Saved-document comparison

32024R1689.txt version record.

This page displays a comparison between saved source copies. It does not state what the source means or determine legal, regulatory, or entity status.

Comparison metadata

Source family	eu-legislation/eurlex-full-text
Check result	source page substantially replaced
Previous saved copy	20260822T222832Z docs_evidence\eu-legislation\eurlex-full-text\20260822T222832Z\32024R1689.txt · sha256 bc6030f7d844db621962657c4203bc7d5432352cca9f5e27b9471f2584fcbb09
Latest saved copy	20260823T093000Z docs_evidence\eu-legislation\eurlex-full-text\20260823T093000Z\32024R1689.txt · sha256 7c5f3cd3af3d7235dd22c8093fcb34ff0c9aa00bbbe693ab537c5a7c046e5cbc
Computed difference	96.77% of lines changed, computed as changed units / max(previous units, latest units) × 100.

Changed saved lines

− REGUL A TION (EU) 2024/1689 OF THE EUR OPEAN P ARLIAMENT AND OF THE CO UNCIL

− laying do wn har monised r ules on ar tif icial intelligence and amending Regulations (EC) No 300/2008,

− (EU) No 167/2013, (EU) No 168/2013, (EU) 2018/858, (EU) 2018/1139 and (EU) 2019/2144 and

− Directiv es 2014/90/EU, (EU) 2016/797 and (EU) 2020/1828 (Ar tif icial Intelligence A ct)

− (T ext with EEA relevance)

− THE EUR OPEAN P ARLIAMENT AND THE COUNCIL OF THE EUR OPEAN UNION,

− Having regard to the T reaty on the Functioning of the European Union, and in par ticular Ar ticles 16 and 114 thereof,

− Af ter transmission of the draf t legislative act to the national parliaments,

− Having regard to the opinion of the European Economic and Social Committe e (

− 1

− ),

− Having regard to the opinion of the European Central Bank (

− 2

− ),

− Having regard to the opinion of the Committee of the Regions (

− 3

− ),

− A cting in accordance with the ordinar y legislative procedure (

− 4

− ),

− (1) The pur pose of this Regulation is to imp rove the functioning of the internal marke t by la ying do wn a unif or m leg al

− framew ork in par ticular f or the development, the placing on the market, the putting into ser vice and the use of

− ar tificial inte lligence syste ms (AI systems) in the Uni on, in accordance with Union values, to promote the uptak e of

− human centr ic and tr ustwor thy ar tificial intellig ence (AI) while ensur ing a high level of prot ection of health, saf ety ,

− fundamental r ights as enshr ined in the Char te r of Fundamental Rights of the European Union (the ‘Char te r ’),

− including democracy , the r ule of law and environmental prot ection, to protect against the har mful effe cts of AI

− syste ms in the Uni on, and to suppor t inno vation. This Regulation ensures the free moveme nt, cross-border , of

− AI-based goods and ser vices, thus preventing Member Stat es from imp osing restr ictions on the development,

− mark eting and use of AI systems, unless explicitly author ised by this Regulation.

− (2) This Regulation should be applied in accordance with the values of the Uni on enshr ined as in the Char te r , f acilitating

− the protection of natural persons, under takings, democracy , the r ule of law and environmental prot ection, while

− boosting innovation and emplo yment and making the Uni on a leader in the up take of tr ustwo r th y AI.

− (3) AI systems can be easily deplo yed in a larg e var iety of sect ors of the economy and many par ts of society , including

− across borders, and can easily circulat e throughout the Uni on. Cer tain Member States hav e already explored the

− adop tion of national r ules to ensure that AI is tr ustwor th y and saf e and is developed and used in accordance with

− fundamental r ights obliga tions. Diverging national r ules ma y lead to the fragmentation of the inter nal market and

− ma y decrease leg al cer tainty f or operators that develop, im por t or use AI systems. A consistent and high level of

− prot ection throughout the Union should theref ore be ensured in order to achi eve tr ustw or th y AI, while divergence s

− ham per ing the free circulation, inno vation, deplo yment and the uptak e of AI systems and related products and

− ser vices within the inter nal marke t should be prevent ed by la ying do wn unif or m obligations f or operato rs and

− Offi cial Jour nal

− of the European Union

− EN

− L ser ies

− 2024/1689

− 12.7.2024

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 1/144

− (

− 1

− ) OJ C 517, 22.12.2021, p. 56.

− (

− 2

− ) OJ C 115, 11.3.2022, p. 5.

− (

− 3

− ) OJ C 97, 28.2.2022, p. 60.

− (

− 4

− ) P osition of the European Parl iament of 13 March 2024 (not yet published in the Official Jour nal) and decision of the Council of

− 21 Ma y 2024.

− guarante eing the unif or m protect ion of ove r r iding reasons of public inte rest and of r ights of persons throughout the

− inte r nal marke t on the basis of Ar ticle 114 of the T reaty on the Functioning of the European Union (TFEU). T o the

− exte nt that this Regulation contains specif ic r ules on the protect ion of individuals with regard to the processing of

− personal data concer ning restr ictions of the use of AI syste ms f or remote biometr ic identification f or the pur pose of

− la w enf orcement, of the use of AI systems f or r isk assessments of natural persons f or the pur pose of la w

− enf orcement and of the use of AI systems of biometr ic cate gor isation f or the pur pose of law enf orcement, it is

− appropr iat e to base this Regulation, in so f ar as those specific r ules are concer ned, on Ar ticle 16 TFEU. In light of

− those specific r ules and the recourse to Ar ticle 16 TFEU, it is appropr iate to consult the European Data Protection

− Board.

− (4) AI is a f ast ev olving f amily of t echnologies that contr ibut es to a wide ar ra y of economic, environmental and societal

− benefits across the entire spectr um of industr ies and social activities. By imp roving prediction, optimising operations

− and resource allocation, and personalising digital solutions ava ilable f or individuals and org anisations, the use of AI

− can provide k ey comp etitive advantages t o under takings and suppor t socially and environmentally beneficial

− outcomes, f or exam ple in healthcare, agr iculture, f ood safety , education and training, media, spor ts, culture,

− infrastr ucture manag ement, energy , transpor t and logistics, public ser vices, secur ity , justice, resource and energy

− efficiency , en vironmental monitoring, the conser vation and restoration of biodiversity and ecosystems and climate

− ch ange mitig ation and adap tation.

− (5) At the same time, depending on the circumstances regarding its specific application, use, and level of te chnological

− development, AI ma y g enerate r isks and cause har m to public interests and fundamental r ights that are protect ed by

− Uni on la w . Such har m might be material or immater ial, including physical, psy ch ological, societal or economic

− har m.

− (6) Given the major im pact that AI can hav e on society and the need to build tr ust, it is vital f or AI and its regulator y

− framew ork to be developed in accordance with Union values as enshr ined in Ar ticle 2 of the T reaty on European

− Uni on (TEU), the fundamental r ights and freedoms enshr ined in the T reaties and, pursuant to Ar ticle 6 TEU, the

− Char te r . As a prerequisite , AI should be a human-centr ic tec hnology . It should ser ve as a too l f or people, with the

− ultimate aim of increasing human well-being.

− (7) In order to ensure a consistent and high level of protection of public intere sts as regard s health, saf ety and

− fundamental r ights, common r ules f or high-r isk AI syste ms should be established. Those r ules should be consiste nt

− with the Char te r , non-discr iminat or y and in line with the Uni on’s inte r national trade commitments. They should

− also take into account the European Declaration on Digital Rights and Pr inciples f or the Digital Decade and the

− Ethics guidelines f or tr ustw or th y AI of the High-Level Exper t Group on Ar tif icial Intellig ence (AI HLEG).

− (8) A Union lega l framew ork la ying do wn har monised r ules on AI is theref ore needed to f oster the development, use

− and up take of AI in the inter nal marke t that at the same time meets a high level of protect ion of public interests, such

− as health and saf ety and the prot ection of fundamental r ights, including democracy , the r ule of law and

− en vironmental protect ion as recognised and protect ed by Uni on la w . T o ac hieve that objective, r ules regulating the

− placing on the marke t, the putting into ser vice and the use of cer tain AI systems should be laid down, thus ensur ing

− the smooth functioning of the inte r nal market and allowi ng those systems to benefit from the pr inciple of free

− move ment of goods and ser vices. Those r ules should be clear and robust in protecting fundamental r ights,

− suppor tive of new innovative solutions, enabling a European ecosyste m of public and pr ivate actor s creating AI

− syste ms in line with Union values and unlocking the potential of the digital transf or mation across all regions of the

− Uni on. By la ying down those r ules as well as measures in suppor t of inno vation with a par ticular f ocus on small and

− medium ent er pr ises (SMEs), including star tups, this Regulation suppor ts the objective of promoting the European

− human-centr ic approach to AI and being a g lobal leader in the development of secure, tr ustwor thy and ethical AI as

− stat ed by the European Council (

− 5

− ), and it ensures the prot ection of ethical pr inciples, as specific ally request ed by the

− European Pa rliament (

− 6

− ).

− 2/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− (

− 5

− ) European Council, Special meeting of the European Council (1 and 2 October 2020) — Conclusions, EUC O 13/20, 2020, p. 6.

− (

− 6

− ) European Parl iament resolution of 20 October 2020 with recommendations to the Commission on a framew ork of ethical aspects

− of ar tificia l intellig ence, robotics and relat ed tec hnologies, 2020/2012(INL).

− (9) Har monised r ules applicable to the placing on the mark et, the putting into ser vice and the use of high-r isk AI

− syste ms should be laid do wn consistently with Regulation (EC) No 765/2008 of the European Pa rliament and of the

− Council (

− 7

− ), Decision No 768/2008/EC of the European Pa rliament and of the Council (

− 8

− ) and Regulation (EU)

− 2019/1020 of the European P arliament and of the Council (

− 9

− ) (New Legislative Framewo rk). The har monised r ules

− laid down in this Regulation should apply across sectors and, in line with the New Legislative Framework, should be

− without prejudice to existing Uni on law , in par ticular on data protect ion, consumer prot ection, fundamental r ights,

− em plo yment, and prot ection of work ers, and product safety , to which this Regulation is comp lementar y . As

− a consequence, all r ights and remedies provid ed f or by suc h Union law t o consumers, and other persons on whom

− AI syste ms ma y hav e a nega tive impact, including as rega rds the comp ensation of possible damage s pursuant to

− Council Directive 85/374/EEC (

− 10

− ) remain unaff ected and fully applicable. Fur ther more, in the context of

− em plo yment and protect ion of wo rkers, this Regulation should theref ore not affe ct Uni on law on social policy and

− national labour law , in compliance with Uni on law , concer ning emplo yment and wo rking conditions, including

− health and saf ety at work and the relationship between emp lo yers and w orkers. This Regulation should also not

− aff ect the ex ercise of fundamental r ights as recognised in the Member Stat es and at Uni on level, including the r ight or

− freedom to str ike or to take other action covered b y the specif ic industr ial relations syste ms in Member States as well

− as the r ight to negotiat e, to conclude and enf orce collective agreements or t o take collective action in accordance

− with national law . This Regulation should not affect the provisions aiming t o im prove working conditions in

− platf or m wo rk laid down in a Directive of the European Parliament and of the Council on imp roving working

− conditions in platf or m work. Moreover , this Regulation aims to strengthen the effe ctiveness of suc h existing r ights

− and remedies by establishing specific requirements and obliga tions, including in respect of the transparency ,

− t echnical documentation and record-keepi ng of AI syste ms. Fur ther more, the obliga tions placed on var ious

− operat ors in volved in the AI value chain under this Regulation should apply without prejudice to national la w , in

− com pliance with Uni on law , having the eff ect of limiting the use of cer tain AI systems where suc h la w f alls outside

− the scope of this Regulation or pursues legitimate public interest objectives other than those pursued by this

− Regulation. For exam ple, national labour law and la w on the protection of minors, namely persons below the age of

− 18, taking into account the UNCR C General Comment No 25 (2021) on children’s r ights in relation to the digital

− en vironment, insofa r as the y are not specif ic to AI syste ms and pursue other legitimate public inte rest objectives,

− should not be affe cted by this Regulation.

− (10) The fundamental r ight to the prot ection of personal data is safegua rded in par ticular by Regulations (EU)

− 2016/679 (

− 11

− ) and (EU) 2018/1725 (

− 12

− ) of the European Pa rliament and of the Council and Directive (EU) 2016/680

− of the European Parliament and of the Council (

− 13

− ). Directive 2002/58/EC of the European Parliament and of the

− Council (

− 14

− ) additionally prot ects pr ivate lif e and the confi dentiality of communications, including by wa y of

− pro viding conditions f or any st or ing of personal and non-personal data in, and access from, ter minal equipment.

− Those Union lega l acts provide the basis f or sustainable and responsible data processing, including where data sets

− include a mix of personal and non-personal data. This Regulation does not seek to affect the application of existing

− Uni on law gover ning the processing of personal data, including the tasks and po wers of the independent super visor y

− author ities compet ent to monito r compliance with those instr uments. It also does not affe ct the obliga tions of

− pro viders and deplo yers of AI syste ms in their role as data controllers or processors st emming from Union or

− national law on the prot ection of personal data in so f ar as the design, the development or the use of AI syste ms

− in v olves the processing of personal data. It is also appropr iate to clar ify that data subjects continue t o enjo y all the

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 3/144

− (

− 7

− ) Regulation (EC) No 765/2008 of the European Parl iament and of the Council of 9 July 2008 setting out the requirements f or

− accreditation and repealing Regulation (EEC) No 339/93 (OJ L 218, 13.8.2008, p. 30).

− (

− 8

− ) Decision No 768/2008/EC of the European Parliament and of the Council of 9 July 2008 on a common framework f or the

− mark eting of products, and repealing Council Decision 93/465/EEC (OJ L 218, 13.8.2008, p. 82).

− (

− 9

− ) Regulation (EU) 2019/1020 of the European Parl iament and of the Council of 20 June 2019 on market sur veillance and compliance

− of products and amending Directive 2004/42/EC and Regulations (EC) No 765/2008 and (EU) No 305/2011 (OJ L 169, 25.6.2019,

− p. 1).

− (

− 10

− ) Council Directive 85/374/EEC of 25 July 1985 on the appro ximation of the la ws, regulations and administrative provisions of the

− Member States concer ning liability f or defe ctive products (OJ L 210, 7.8.1985, p. 29).

− (

− 11

− ) Regulation (EU) 2016/679 of the European Parliament and of the Council of 27 Apr il 2016 on the prot ection of natural persons

− with regard t o the processing of personal data and on the free mo vement of suc h data, and repealing Directive 95/46/EC (General

− Data Protection Regulation) (OJ L 119, 4.5.2016, p. 1).

− (

− 12

− ) Regulation (EU) 2018/1725 of the European Parliament and of the Council of 23 October 2018 on the protection of natural

− persons with regard t o the processing of personal data b y the Union institutions, bodies, offices and agencies and on the free

− mo vement of such data, and repealing Regulation (EC) No 45/2001 and Decision No 1247/2002/EC (OJ L 295, 21.11.2018, p. 39).

− (

− 13

− ) Directive (EU) 2016/680 of the European Parl iament and of the Council of 27 Apr il 2016 on the prot ection of natural persons with

− reg ard t o the processing of personal data b y compet ent author ities f or the pur poses of the prevention, investig ation, detection or

− prosecution of cr iminal offences or the execution of cr iminal penalties, and on the free movement of such data, and repealing

− Council Framework Decision 2008/977/JHA (OJ L 119, 4.5.2016, p. 89).

− (

− 14

− ) Directive 2002/58/EC of the European Parl iament and of the Council of 12 July 2002 concer ning the processing of personal data

− and the prot ection of pr ivacy in the electronic communications sector (Directive on pr ivacy and electronic communications) (OJ

− L 201, 31.7.2002, p. 37).

− r ights and guarantees a warded to them b y such Uni on law , including the r ights related to solely automat ed individual

− decision-making, including prof iling. Har monised r ules f or the placing on the marke t, the putting into ser vice and

− the use of AI systems established under this Regulation should f acilitate the effe ctive imp lementation and enable the

− ex ercise of the data subjects’ r ights and other remedies guaranteed under Uni on law on the protect ion of personal

− data and of other fundamental r ights.

− (11) This Regulation should be without prejudice to the provisions rega rding the liability of provider s of inte r mediar y

− ser vices as set out in Regulation (EU) 2022/2065 of the European Parliament and of the Council (

− 15

− ).

− (12) The notion of ‘ AI system’ in this Regulation should be clearly defined and should be closely aligned with the w ork of

− inte r national organisations w orking on AI to ensure leg al cer tainty , f acilitate inter national conve rgence and wide

− accep tance, while providing the f lexibility to accommodate the rapid te chnologi cal developments in this field.

− Moreo ver , the def inition should be based on k ey char acter istics of AI systems that distinguish it from simpler

− traditional sof tware syste ms or programming approac hes and should not cover syste ms that are based on the r ules

− defined solely by natural persons to auto matically ex ecute operations. A ke y ch aracter istic of AI syste ms is their

− capability to infer . This capability t o infer refers t o the process of obtaining the outputs, suc h as predictions, cont ent,

− recommendations, or decisions, which can inf luence physica l and vir tual envir onments, and to a capability of AI

− syste ms to der ive models or algor ithms, or both, from in puts or data. The tec hniques that enable inf erence while

− building an AI syste m include machi ne lear ning approaches that lear n from data how t o ac hieve cer tain objectives,

− and logic- and kno wledge-based approac hes that inf er from encoded knowle dge or symbolic representation of the

− task to be solved. The capacity of an AI syste m to infer transcends basic data processing by enabling lear ning,

− reasoning or modelling. The ter m ‘mach ine-based’ refe rs to the f act that AI systems r un on machi nes. The reference

− t o explicit or im plicit objectives underscores that AI syste ms can operat e according to explicit def ined objectives or

− t o imp licit objectives. The objectives of the AI system ma y be diff erent from the inte nded pur pose of the AI system

− in a specific cont ext. For the pur poses of this Regulation, envir onments should be understood to be the contexts in

− which the AI systems operate, whereas outputs g enerated by the AI syste m ref lect diffe rent functions perfo r med by

− AI syste ms and include predictions, cont ent, recommendations or decisions. AI syste ms are designed to operate with

− var ying levels of autonom y , meaning that the y hav e some degree of independence of actions from human

− in v olvement and of capabilities to operat e without human inte r vention. The adap tiveness that an AI syste m could

− exhibit af te r deplo yment, refe rs to self-lear ning capabilities, allowi ng the syste m to ch ange while in use. AI syste ms

− can be used on a stand-alone basis or as a compo nent of a product, ir respective of whether the syste m is phys ically

− inte grated into the product (embedded) or ser ves the functionality of the product without being integrat ed therein

− (non-embedded).

− (13) The notion of ‘deplo y er ’ refe r red t o in this Regulation should be inte r preted as any natural or lega l person, including

− a public author ity , agency or other body , using an AI system under its author ity , excep t where the AI syste m is used

− in the course of a personal non-profess ional activity . Depending on the type of AI syste m, the use of the system ma y

− aff ect persons other than the deplo y er .

− (14) The notion of ‘biometr ic data’ used in this Regulation should be inter preted in light of the notion of biometr ic data

− as defined in Ar ticle 4, point (14) of Regulation (EU) 2016/679, Ar ticle 3, point (18) of Regulation (EU) 2018/1725

− and Ar ticle 3, point (13) of Directive (EU) 2016/680. Biometr ic data can allo w f or the authentication, identification

− or cate gor isation of natural persons and f or the recognition of emotions of natural persons.

− (15) The notion of ‘biometr ic identifi cation’ refe r red to in this Regulation should be def ined as the auto mated recognition

− of physical, physiological and behavi oural human features such as the f ace, ey e move ment, body shape, v oice,

− prosody , g ait, posture, hear t rate, blood pressure, odour , k eystrok es ch aracter istics, f or the pur pose of establishing an

− individual’s identity b y compari ng biometr ic data of that individual to stored biometr ic data of individuals in

− a refere nce database, ir respective of whether the individual has given its consent or not. This ex cludes AI syste ms

− inte nded to be used f or biometr ic ver ificati on, which includes authentication, whose sole pur pose is to confi r m that

− a specific natural person is the person he or she claims to be and to confi r m the identity of a natural person f or the

− sole pur pose of hav ing access to a ser vice, unlocking a device or hav ing secur ity access to premises.

− 4/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− (

− 15

− ) Regulation (EU) 2022/2065 of the European Parliament and of the Council of 19 October 2022 on a Sing le Market For Digital

− Ser vices and amending Directive 2000/31/EC (Digital Ser vices A ct) (OJ L 277, 27.10.2022, p. 1).

− (16) The notion of ‘biometr ic catego r isation’ refer red to in this Regulation should be def ined as assigning natural persons

− t o specif ic cate gor ies on the basis of their biometr ic data. Such specif ic catego r ies can relate to aspects suc h as sex,

− ag e, hair colour , e ye colour , tattoos, behavio ural or personality traits, languag e, religion, membership of a national

− minor ity , sexual or political or ientation. This does not include biometr ic catego r isation systems that are a purely

− ancillar y f eature intr insically linked to another commercial ser vice, meaning that the f eature cannot, f or objective

− t echnical reasons, be used without the pr incipal ser vice, and the inte gration of that feature or functionality is not

− a means to circum vent the applicability of the r ules of this Regulation. For example, filters cate gor ising f acial or body

− f eatures used on online marke tplaces could constitute suc h an ancillar y f eature as they can be used only in relation to

− the pr incipal ser vice which consists in selling a product by allowing the consumer to preview the displa y of the

− product on him or herself and help the consumer to mak e a purcha se decision. Filte rs used on online social netw ork

− ser vices which cate gor ise f acial or body f eatures t o allow users to add or modify pictures or videos could also be

− considered to be ancillar y f eature as suc h filter cannot be used without the pr incipal ser vice of the social netw ork

− ser vices consisting in the shar ing of cont ent online.

− (17) The notion of ‘remot e biometr ic identifica tion syste m’ refe r red to in this Regulation should be def ined functionally ,

− as an AI syste m inte nded f or the identifica tion of natural persons without their active in volvement, typically at

− a distance, through the comp ar ison of a person’s biometr ic data with the biometr ic data contained in a reference

− database, ir respectively of the par ticular tec hnology , processes or types of biometr ic data used. Such remote

− biometr ic identifica tion systems are typically used to perceive multiple persons or their behavio ur simultaneously in

− order t o f acilitate signif icantly the identification of natural persons without their active inv olvement. This ex cludes

− AI syste ms inte nded t o be used f or biometr ic ver if ication, which includes authentication, the sole pur pose of which

− is to confi r m that a specif ic natural person is the person he or she claims to be and to confi r m the identity of

− a natural person f or the sole pur pose of hav ing access to a ser vice, unloc king a device or hav ing secur ity access to

− premises. That exclusion is justified b y the f act that suc h syste ms are like ly to hav e a minor imp act on fundamental

− r ights of natural persons comp ared to the remote biometr ic identifi cation syste ms which ma y be used f or the

− processing of the biometr ic data of a larg e number of persons without their active in v olvement. In the case of

− ‘real-time’ syste ms, the captu r ing of the biometr ic data, the comp ar ison and the identifi cation occur all

− instantaneously , near -instantaneously or in any event without a significant dela y . In this regard, there should be no

− scope f or circum venting the r ules of this Regulation on the ‘real-time’ use of the AI systems concer ned by providing

− f or minor dela ys. ‘Real-time’ syste ms in v olve the use of ‘live’ or ‘near -live’ material, suc h as video f ootage , generat ed

− b y a camera or other device with similar functionality . In the case of ‘post’ systems, in contrast, the biometr ic data

− has already been captured and the comp ar ison and identifica tion occur only af ter a significant dela y . This inv olves

− mat er ial, such as pictures or video f ootage g enerated by closed circuit te levision cameras or pr ivate devices, which

− has been g enerated bef ore the use of the syste m in respect of the natural persons concer ned.

− (18) The notion of ‘emotion recognition system’ refe r red to in this Regulation should be def ined as an AI syste m f or the

− pur pose of identifying or inferr ing emotions or intentions of natural persons on the basis of their biometr ic data.

− The notion refe rs to emotions or intent ions suc h as happiness, sadness, anger , sur pr ise, disgust, embar rassment,

− ex cite ment, shame, contem pt , satisfaction and amusement. It does not include physical stat es, such as pain or

− f atigue, including, f or exam ple, syste ms used in detecting the state of f atigue of profe ssional pilots or dr ivers f or the

− pur pose of preventing accidents. This does also not include the mere detection of readily apparent expressions,

− g estures or movements, unless the y are used f or identifying or inferr ing emotions. Those expressions can be basic

− f acial expressions, suc h as a frown or a smile, or gestures suc h as the moveme nt of hands, ar ms or head, or

− ch aracteristics of a person’s voice, suc h as a raised voice or whisper ing.

− (19) For the pur poses of this Regulation the notion of ‘publicly accessible space’ should be understood as refe r r ing to any

− phys ical space that is accessible to an undeter mined number of natural persons, and ir respective of whether the

− space in question is pr ivat ely or publicly ow ned, ir respective of the activity f or which the space ma y be used, suc h as

− f or commerce, f or exam ple, shops, restaurants, cafés; f or ser vices, f or exam ple, banks, professional activities,

− hospitality ; f or spor t, f or exam ple, swimming pools, gyms, stadiums; f or transpor t, f or exam ple, bus, metro and

− railwa y stations, air por ts, means of transpor t ; f or ent er tainment, f or exam ple, cinemas, theatres, museums, concer t

− and confe rence halls; or f or leisure or other wise, f or example, public roads and squares, park s, f orests, pla ygrounds.

− A space should also be classified as being publicly accessible if, regard less of pote ntial capacity or secur ity

− restr ictions, access is subject t o cer tain predete r mined conditions which can be fulf illed by an undeter mined number

− of persons, suc h as the purc hase of a tick et or title of transpor t, pr ior registration or hav ing a cer tain age. In contrast,

− a space should not be considered to be publicly accessible if access is limited to specif ic and def ined natural persons

− through either Uni on or national law directly related t o public safety or secur ity or through the clear manif estation

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 5/144

− of will by the person having the relevant author ity ove r the space. The f actual possibility of access alone, such as an

− unloc ked door or an open g at e in a f ence, does not imp ly that the space is publicly accessible in the presence of

− indications or circumstances suggesting the contrar y , suc h as. signs prohibiting or restr icting access. Compan y and

− f actory premises, as well as offices and workplaces that are intende d to be accessed only by relevant emp lo yees and

− ser vice providers, are spaces that are not publicly accessible. Publicly accessible spaces should not include pr isons or

− border control. Some other spaces ma y compr ise both publicly accessible and non-publicly accessible spaces, suc h as

− the hallwa y of a pr ivat e residential building necessar y to access a docto r ’s off ice or an air por t. Online spaces are not

− co vered, as they are not physical spaces. Whether a given space is accessible to the public should howe ver be

− det er mined on a case-by-c ase basis, having regard to the specif icities of the individual situation at hand.

− (20) In order to obtain the great est benefits from AI syste ms while prot ecting fundamental r ights, health and saf ety and to

− enable democratic control, AI lite racy should equip providers, deplo yers and affect ed persons with the necessar y

− notions to mak e inf or med decisions regarding AI systems. Those notions ma y var y with regard to the relevant

− cont ext and can include understanding the cor rect application of t echnical elements dur ing the AI system’s

− development phase, the measures to be applied dur ing its use, the suitable wa ys in which to inter pret the AI system’s

− output, and, in the case of affe cted persons, the kno wledge necessar y to understand how decisions tak en with the

− assistance of AI will hav e an imp act on them. In the context of the application this Regulation, AI literacy should

− pro vide all relevant actors in the AI value ch ain with the insights required to ensure the appropr iat e compliance and

− its cor rect enf orcement. Fur ther more, the wide im plementation of AI literac y measures and the introduction of

− appropr iat e f ollow-up actions could contr ibut e to impro ving work ing conditions and ultimately sustain the

− consolidation, and inno vation path of tr ustwor thy AI in the Uni on. The European Ar tificial Intellig ence Board (the

− ‘Board’) should suppor t the Commission, to promot e AI litera cy tools, public a wareness and understanding of the

− benefits, r isks , safegua rds, r ights and obliga tions in relation to the use of AI syste ms. In cooperation with the relevant

− stak eholders, the Commission and the Member States should f acilitate the drawing up of v oluntar y codes of conduct

− t o advance AI lite racy among persons dealing with the development, operation and use of AI.

− (21) In order to ensure a level pla ying field and an effective protect ion of r ights and freedoms of individuals across the

− Uni on, the r ules established by this Regulation should apply to provid ers of AI syste ms in a non-discr iminat or y

− manner , ir respective of whether the y are established within the Uni on or in a third countr y , and to deplo y ers of AI

− syste ms established within the Uni on.

− (22) In light of their digital nature, cer tain AI syste ms should f all within the scope of this Regulation even when they are

− not placed on the marke t, put into ser vice, or used in the Uni on. This is the case, f or example, where an operator

− established in the Union contracts cer tain ser vices to an operator established in a third countr y in relation t o an

− activity to be perform ed by an AI system that would qualify as high-r isk. In those circumstances, the AI syste m used

− in a third countr y by the operator could process data law fully collected in and transferre d from the Uni on, and

− pro vide to the contracting operato r in the Uni on the output of that AI system resulting from that processing,

− without that AI syste m being placed on the marke t, put into ser vice or used in the Uni on. T o prevent the

− circum vention of this Regulation and to ensure an effe ctive prot ection of natural persons located in the Union, this

− Regulation should also apply to providers and deplo yers of AI systems that are established in a third countr y , to the

− exte nt the output produced b y those syste ms is intende d t o be used in the Union. Nonetheless, to take into account

− existing ar rang ements and special needs f or future cooperation with f oreign par tners with whom inf or mation and

− evidence is ex chang ed, this Regulation should not apply to public author ities of a third countr y and inter national

− org anisations when acting in the framework of cooperation or inter national agreements concluded at Uni on or

− national level f or law enf orcement and judicial cooperation with the Union or the Member Stat es, provided that the

− relevant third countr y or inte r national organisation provides adequate safegua rds with respect to the prot ection of

− fundamental r ights and freedoms of individuals. Where relevant, this ma y co ver activities of entities entr usted by the

− third countr ies to car r y out specif ic tasks in suppor t of suc h law enf orcement and judicial cooperation. Such

− framew ork f or cooperation or agreements hav e been established bilat erally between Member States and third

− countr ies or between the European Union, Europol and other Union agencies and third countr ies and inte r national

− org anisations. The author ities compet ent f or super vision of the law enf orcement and judicial author ities under this

− Regulation should assess whether those framew orks f or cooperation or inter national agreements include adequate

− saf eguards with respect to the protect ion of fundamental r ights and freedoms of individuals. Recipient national

− 6/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− author ities and Uni on institutions, bodies, off ices and agencies making use of suc h outputs in the Uni on remain

− accountable to ensure their use comp lies with Uni on law . When those inter national agreements are revised or new

− ones are concluded in the future, the contracting par ties should make utmost eff or ts to align those agreements with

− the requirements of this Regulation.

− (23) This Regulation should also apply to Uni on institutions, bodies, off ices and agencies when acting as a provid er or

− deplo y er of an AI syste m.

− (24) If, and insofa r as, AI systems are placed on the marke t, put into ser vice, or used with or without modification of such

− syste ms f or militar y , defe nce or national secur ity pur poses, those should be excluded from the scope of this

− Regulation regard less of which type of entity is car r ying out those activities, such as whether it is a public or pr ivate

− entity . As regard s militar y and defe nce pur poses, suc h ex clusion is justif ied both by Ar ticle 4(2) TEU and b y the

− specificities of the Member Stat es’ and the common Union defence policy covered by Chapt er 2 of Title V TEU that

− are subject to public inter national law , which is theref ore the more appropr iate legal framew ork f or the regulation of

− AI systems in the cont ext of the use of lethal f orce and other AI syste ms in the context of militar y and defence

− activities. As rega rds national secur ity pur poses, the ex clusion is justified both by the fact that national secur ity

− remains the sole responsibility of Member States in accordance with Ar ticle 4(2) TEU and by the specif ic nature and

− operational needs of national secur ity activities and specific national r ules applicable to those activities. Nonetheless,

− if an AI system developed, placed on the marke t, put into ser vice or used f or militar y , defence or national secur ity

− pur poses is used outside those te m porar ily or per manently f or other pur poses, f or exam ple, civilian or humanitar ian

− pur poses, law enf orcement or public secur ity pur poses, suc h a system would f all within the scope of this Regulation.

− In that case, the entity using the AI syste m f or other than militar y , defe nce or national secur ity pur poses should

− ensure the compliance of the AI syste m with this Regulation, unless the syste m is already compliant with this

− Regulation. AI syste ms placed on the marke t or put into ser vice f or an ex cluded pur pose, namely militar y , defe nce or

− national secur ity , and one or more non-exclu ded pur poses, suc h as civilian pur poses or law enf orcement, f all within

− the scope of this Regulation and pro viders of those systems should ensure compliance with this Regulation. In those

− cases, the f act that an AI system ma y f all within the scope of this Regulation should not affe ct the possibility of

− entities car r ying out national secur ity , defence and militar y activities, rega rdless of the type of entity car r ying out

− those activities, to use AI systems f or national secur ity , militar y and defe nce pur poses, the use of which is excluded

− from the scope of this Regulation. An AI system placed on the marke t f or civilian or law enf orcement pur poses

− which is used with or without modifi cation f or militar y , defence or national secur ity pur poses should not fall within

− the scope of this Regulation, rega rdless of the type of entity car r ying out those activities.

− (25) This Regulation should suppor t inno vation, should respect freedom of science, and should not under mine research

− and development activity . It is theref ore necessar y to ex clude from its scope AI syste ms and models specifically

− developed and put into ser vice f or the sole pur pose of scientifi c research and development. Moreove r , it is necessar y

− t o ensure that this Regulation does not other wise affect scientific research and development activity on AI syste ms or

− models pr ior to being placed on the market or put into ser vice. As regards product-or iented research, testing and

− development activity rega rding AI syste ms or models, the provisions of this Regulation should also not apply pr ior

− t o those syste ms and models being put into ser vice or placed on the marke t. That ex clusion is without prejudice to

− the obligation to comp ly with this Regulation where an AI system f alling into the scope of this Regulation is placed

− on the marke t or put into ser vice as a result of suc h research and development activity and to the application of

− pro visions on AI regulato r y sandbo xe s and te sting in real world conditions. Fur ther more, without prejudice to the

− ex clusion of AI syste ms specif ically developed and put into ser vice f or the sole pur pose of scientifi c researc h and

− development, any other AI system that ma y be used f or the conduct of any research and development activity should

− remain subject t o the provisions of this Regulation. In any event, any research and development activity should be

− car r ied out in accordance with recognised ethical and profe ssional standards f or scientific research and should be

− conducte d in accordance with applicable Uni on law .

− (26) In order to introduce a propor tionate and effe ctive set of binding r ules f or AI syste ms, a clearly defined r isk -based

− approac h should be f ollo wed. That approac h should tailor the type and content of suc h r ules t o the inte nsity and

− scope of the r isks that AI syste ms can g enerate. It is theref ore necessar y to prohibit cer tain unaccept able AI practices,

− t o la y do wn requirements f or high-r isk AI systems and obligati ons f or the relevant operators, and to la y do wn

− transparency obligations f or cer tain AI syste ms.

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 7/144

− (27) While the r isk -based approac h is the basis f or a propor tionate and effe ctive set of binding r ules, it is imp or tant to

− recall the 2019 Ethics guidelines f or tr ustwo r th y AI developed by the independent AI HLEG appoint ed by the

− Commission. In those guidelines, the AI HLEG developed seven non-binding ethical pr inciples f or AI which are

− inte nded to help ensure that AI is tr ustwor thy and ethically sound. The seven pr inciples include human agency and

− o versight ; te chnical robustness and safety; pr ivacy and data gover nance; transparency ; diversity , non-discr imination

− and f air ness; societal and envir onmental well-being and accountability . Without prejudice to the leg ally binding

− requirements of this Regulation and any other applicable Uni on law , those guidelines contr ibut e to the design of

− coherent, tr ustwor thy and human-centr ic AI, in line with the Char te r and with the values on which the Uni on is

− f ounded. A ccording to the guidelines of the AI HLEG, human age ncy and oversight means that AI systems are

− developed and used as a t ool that ser ves people, respects human dignity and personal autonom y , and that is

− functioning in a wa y that can be appropr iat ely controlled and ove rseen by humans. T ec hnical robustness and saf ety

− means that AI syste ms are developed and used in a wa y that allows robustness in the case of problems and resilience

− ag ainst atte mp ts to alter the use or perform ance of the AI syste m so as to allow unla wful use by third par ties, and

− minimise uninte nded har m. Pr ivacy and data gove r nance means that AI syste ms are developed and used in

− accordance with pr ivacy and data prot ection r ules, while processing data that meets high standards in ter ms of

− quality and inte gr ity . T ransparency means that AI syste ms are developed and used in a wa y that allo ws appropr iate

− traceability and explainability , while making humans aw are that they communicate or interact with an AI syste m, as

− well as duly inf or ming deplo yers of the capabilities and limitations of that AI system and affe cted persons about their

− r ights. Diversity , non-discr imination and fairne ss means that AI syste ms are developed and used in a wa y that

− includes diverse actor s and promotes equal access, g ender equality and cultural diversity , while av oiding

− discr iminatory imp acts and unfa ir biases that are prohibite d by Union or national law . Social and envir onmental

− well-being means that AI systems are developed and used in a sustainable and environmentally fr iendly manner as

− well as in a wa y to benefi t all human beings, while monitori ng and assessing the long-term imp acts on the

− individual, society and democracy . The application of those pr inciples should be translat ed, when possible, in the

− design and use of AI models. They should in any case ser ve as a basis f or the draf ting of codes of conduct under this

− Regulation. All stakeholders, including industr y , academia, civil society and standardisation org anisations, are

− encourag ed to take into account, as appropr iate, the ethical pr inciples f or the development of voluntary best

− practices and standards.

− (28) Aside from the many beneficial uses of AI, it can also be misused and provide novel and po werful to ols f or

− manipulative, exploitative and social control practices. Such practices are par ticularly har mful and abusive and

− should be prohibite d because they contradict Union values of respect f or human dignity , freedom, equality ,

− democracy and the r ule of law and fundamental r ights enshr ined in the Char te r , including the r ight to

− non-discr imination, to data prot ection and t o pr ivacy and the r ights of the child.

− (29) AI-enabled manipulative t echniques can be used to persuade persons to eng age in unwant ed behaviours, or to

− deceive them by nudging them into decisions in a wa y that subver ts and imp airs their autonom y , decision-making

− and free choices. The placing on the marke t, the putting into ser vice or the use of cer tain AI syste ms with the

− objective to or the effe ct of mat er ially distor ting human behavio ur , whereb y signifi cant har ms, in par ticular having

− sufficiently imp or tant adverse im pacts on phys ical, psy chological health or financ ial intere sts are like ly to occur , are

− par ticularly dangerous and should theref ore be prohibite d. Such AI syste ms deplo y subliminal comp onents suc h as

− audio, image, video stimuli that persons cannot perceive, as those stimuli are bey ond human percep tion, or other

− manipulative or decept ive tec hniques that subver t or imp air person’s auto nomy , decision-making or free ch oice in

− wa ys that people are not consciously aw are of those te chniques or , where they are aw are of them, can still be

− deceived or are not able to control or resist them. This could be f acilitated, f or example, by machine-brain inte rfaces

− or vir tual reality as they allow f or a higher degree of control of what stimuli are presented t o persons, insofa r as they

− ma y materially distor t their behavio ur in a significantly har mful manner . In addition, AI syste ms ma y also other wise

− exploit the vulnerabilities of a person or a specif ic group of persons due to their age , disability within the meaning of

− Directive (EU) 2019/882 of the European Pa rliament and of the Council (

− 16

− ), or a specific social or economic

− situation that is likely to mak e those persons more vulnerable t o exploitation suc h as persons living in extreme

− po ver ty , ethnic or religious minor ities. Such AI syste ms can be placed on the marke t, put into ser vice or used with

− the objective to or the effe ct of materially distor ting the behavi our of a person and in a manner that causes or is

− reasonably like ly to cause signifi cant har m to that or another person or groups of persons, including har ms that ma y

− be accumulated over time and should theref ore be prohibite d. It ma y not be possible to assume that there is an

− 8/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− (

− 16

− ) Directive (EU) 2019/882 of the European Parliament and of the Council of 17 Apr il 2019 on the accessibility requirements f or

− products and ser vices (OJ L 151, 7.6.2019, p. 70).

− inte ntion to distor t behavio ur where the distor tion results from f actors exter nal to the AI syste m which are outside

− the control of the provid er or the deplo y er , namely f actors that ma y not be reasonably f oreseeable and theref ore not

− possible f or the provider or the deplo y er of the AI syste m to mitigat e. In any case, it is not necessar y f or the provid er

− or the deplo y er t o hav e the inte ntion t o cause signifi cant har m, provided that such har m results from the

− manipulative or exploitative AI-enabled practices. The prohibitions f or suc h AI practices are complement ar y to the

− pro visions contained in Directive 2005/29/EC of the European Parliame nt and of the Council (

− 17

− ), in par ticular unfa ir

− commercial practices leading to economic or financial har ms t o consumers are prohibited under all circumstances,

− ir respective of whether they are put in place through AI systems or other wise. The prohibitions of manipulative and

− exploitative practices in this Regulation should not affe ct lawful practices in the context of medical treatment such as

− psy chologi cal treatment of a mental disease or phys ical rehabilitation, when those practices are car r ied out in

− accordance with the applicable la w and medical standards, f or exam ple explicit consent of the individuals or their

− leg al representatives. In addition, common and legitimate commercial practices, f or example in the fi eld of

− adver tising, that comply with the applicable law should not, in themselves, be rega rded as constituting har mful

− manipulative AI-enabled practices.

− (30) Biometr ic categor isation syste ms that are based on natural persons’ biometr ic data, such as an individual person’s

− f ace or finger pr int, to deduce or inf er an individuals ’ political opinions, trade union membership, religious or

− philosophical beliefs, race, sex life or sexual or ientation should be prohibite d. That prohibition should not cover the

− la wful labelling, filter ing or categori sation of biometr ic data sets acquired in line with Uni on or national la w

− according to biometr ic data, suc h as the sor ting of imag es according t o hair colour or e ye colour , which can f or

− exam ple be used in the area of law enf orcement.

− (31) AI systems provid ing social scor ing of natural persons b y public or pr ivat e actors ma y lead to discr iminat or y

− outcomes and the ex clusion of cer tain groups. They ma y violate the r ight to dignity and non-discr imination and the

− values of equality and justice. Such AI syste ms evaluate or classify natural persons or groups thereof on the basis of

− multiple data points related to their social behavio ur in multiple contexts or known, inf er red or predicted personal

− or personality ch aracter istics ove r cer tain per iods of time. The social score obtained from suc h AI systems ma y lead

− t o the detr imental or unfa v ourable treatment of natural persons or whole groups thereof in social cont exts, which

− are unrelate d to the context in which the data was or iginally generat ed or collect ed or t o a detr imental treatment that

− is dispropor tionate or unjustifie d to the gra vity of their social behavio ur . AI systems entailing suc h unaccept able

− scor ing practices and leading to suc h detr imental or unfa v ourable outcomes should theref ore be prohibited. That

− prohibition should not affe ct lawful evaluation practices of natural persons that are car r ied out f or a specific pur pose

− in accordance with Uni on and national law .

− (32) The use of AI syste ms f or ‘real-time’ remote biometr ic identification of natural persons in publicly accessible spaces

− f or the pur pose of law enf orcement is par ticularly intr usive to the r ights and freedoms of the concer ned persons, to

− the extent that it ma y affect the pr ivate life of a large par t of the population, evok e a f eeling of constant sur veillance

− and indirectly dissuade the exercise of the freedom of assembly and other fundamental r ights. T echnical inaccuracies

− of AI syste ms intende d f or the remote biometr ic identifi cation of natural persons can lead to biased results and entail

− discr iminatory eff ects. Such possible biased results and discr iminatory eff ects are par ticularly relevant with regard to

− ag e, ethnicity , race, sex or disabilities. In addition, the immediacy of the impact and the limited oppor tunities f or

− fur ther chec ks or cor rections in relation to the use of suc h systems operating in real-time car r y height ened r isks f or

− the r ights and freedoms of the persons concer ned in the cont ext of, or im pacted by , law enf orcement activities.

− (33) The use of those systems f or the pur pose of law enf orcement should theref ore be prohibited, ex cept in exhaustively

− list ed and nar ro wly defined situations, where the use is str ictly necessar y to ac hieve a substantial public interest, the

− im por tance of which outweighs the r isks . Those situations inv olve the search f or cer tain victims of cr ime including

− missing persons; cer tain threats to the life or to the phys ical safety of natural persons or of a ter ror ist attac k; and the

− localisation or identification of per petrat ors or suspects of the cr iminal offences listed in an annex to this Regulation,

− where those cr iminal offe nces are punishable in the Member Stat e concer ned by a custo dial sentence or a detention

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 9/144

− (

− 17

− ) Directive 2005/29/EC of the European Parl iament and of the Council of 11 Ma y 2005 concer ning unf air business-to-consumer

− commercial practices in the inter nal market and amending Council Directive 84/450/EEC, Directives 97/7/EC, 98/27/EC and

− 2002/65/EC of the European Parl iament and of the Council and Regulation (EC) No 2006/2004 of the European Parl iament and of

− the Council (‘Unf air Commercial Practices Directive’) (OJ L 149, 11.6.2005, p. 22).

− order f or a maximum per iod of at least f our y ears and as they are defined in the law of that Member Stat e. Such

− a threshold f or the custo dial sent ence or detention order in accordance with national la w contr ibut es to ensur ing

− that the offence should be ser ious enough to potentially justify the use of ‘real-time’ remot e biometr ic identification

− syste ms. Moreover , the list of cr iminal offe nces provided in an annex to this Regulation is based on the 32 cr iminal

− offe nces listed in the Council Framework Decision 2002/584/JHA (

− 18

− ), taking into account that some of those

− offe nces are, in practice, likel y to be more relevant than others, in that the recourse to ‘real-time’ remote biometr ic

− identification could, f oreseeably , be necessar y and propor tionate to highly var ying degrees f or the practical pursuit

− of the localisation or identification of a per petrat or or suspect of the diff erent cr iminal offences listed and having

− rega rd to the like ly differences in the ser iousness, probability and scale of the har m or possible negati ve

− consequences. An imminent threat to life or the physical safety of natural persons could also result from a ser ious

− disr up tion of cr itical infrastr ucture, as def ined in Ar ticle 2, point (4) of Directive (EU) 2022/2557 of the European

− P arliament and of the Council (

− 19

− ), where the disr up tion or destr uction of such cr itical infrastr ucture wo uld result in

− an imminent threat t o life or the physical safety of a person, including through ser ious har m to the provision of basic

− supplies to the population or to the ex ercise of the core function of the State. In addition, this Regulation should

− preser ve the ability f or law enf orcement, border control, immigration or asylum author ities to car r y out identity

− ch ec k s in the presence of the person concer ned in accordance with the conditions set out in Union and national law

− f or such chec ks. In par ticular , law enf orcement, border control, immigration or asylum author ities should be able to

− use inf or mation syste ms, in accordance with Union or national law , to identify persons who, dur ing an identity

− ch ec k , either refuse to be identified or are unable to stat e or prove their identity , without being required b y this

− Regulation t o obtain pr ior author isation. This could be, f or exam ple, a person inv olved in a cr ime, being unwilling,

− or unable due to an accident or a medical condition, to disclose their identity t o law enf orcement author ities.

− (34) In order to ensure that those syste ms are used in a responsible and propor tionate manner , it is also impor tant to

− establish that, in each of those exhaustively list ed and nar ro wly def ined situations, cer tain elements should be take n

− into account, in par ticular as regards the nature of the situation giving r ise to the request and the consequences of

− the use f or the r ights and freedoms of all persons concer ned and the saf eguards and conditions provid ed f or with the

− use. In addition, the use of ‘real-time’ remot e biometr ic identifi cation systems in publicly accessible spaces f or the

− pur pose of la w enf orcement should be deplo y ed only to confir m the specif ically targe ted individual’s identity and

− should be limit ed to what is str ictly necessar y concer ning the per iod of time, as well as the g eographic and personal

− scope, hav ing rega rd in par ticular to the evidence or indications rega rding the threats, the victims or per petrat or . The

− use of the real-time remote biometr ic identifica tion system in publicly accessible spaces should be author ised only if

− the relevant law enf orcement author ity has comp let ed a fundamental r ights im pact assessment and, unless provid ed

− other wise in this Regulation, has register ed the syste m in the database as set out in this Regulation. The reference

− database of persons should be appropr iate f or each use case in each of the situations mentioned above.

− (35) Each use of a ‘real-time’ remote biometr ic identification system in publicly accessible spaces f or the pur pose of law

− enf orcement should be subject to an express and specif ic author isation by a judicial author ity or by an independent

− administrative author ity of a Member State whose decision is binding. Such author isation should, in pr inciple, be

− obtained pr ior to the use of the AI system with a view to identifying a person or persons. Ex ceptions to that r ule

− should be allowed in duly justifi ed situations on grounds of urg ency , namely in situations where the need to use the

− syste ms concer ned is suc h as to mak e it effectively and objectively impossible to obtain an author isation bef ore

− commencing the use of the AI system. In suc h situations of urg ency , the use of the AI syste m should be restr icted to

− the absolute minimum necessar y and should be subject to appropr iate safeguards and conditions, as det er mined in

− national law and specif ied in the cont ext of each individual urgent use case by the law enf orcement author ity itself.

− In addition, the law enf orcement author ity should in such situations request suc h author isation while provid ing the

− reasons f or not hav ing been able t o request it earlier , without undue dela y and at the lat est within 24 hours. If such

− an author isation is rejected, the use of real-time biometr ic identification systems linked to that author isation should

− cease with immediate effect and all the data related to suc h use should be discarded and delete d. Such data includes

− 10/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− (

− 18

− ) Council Framework Decision 2002/584/JHA of 13 June 2002 on the European ar rest war rant and the sur render procedures

− between Member States (OJ L 190, 18.7.2002, p. 1).

− (

− 19

− ) Directive (EU) 2022/2557 of the European Parl iament and of the Council of 14 December 2022 on the resilience of cr itical entities

− and repealing Council Directive 2008/114/EC (OJ L 333, 27.12.2022, p. 164).

− in put data directly acquired b y an AI syste m in the course of the use of such system as well as the results and outputs

− of the use linked to that author isation. It should not include in put that is lega lly acquired in accordance with another

− Uni on or national law . In any case, no decision producing an adverse lega l effe ct on a person should be taken based

− solely on the output of the remot e biometr ic identification system.

− (36) In order to car r y out their tasks in accordance with the requirements set out in this Regulation as well as in national

− r ules, the relevant marke t sur veillance author ity and the national data prot ection author ity should be notif ied of each

− use of the real-time biometr ic identification system. Market sur veillance author ities and the national data prot ection

− author ities that hav e been notif ied should submit t o the Commission an annual repor t on the use of real-time

− biometr ic identifi cation syste ms.

− (37) Fur ther more, it is appropr iate to provid e, within the exhaustive framework set by this Regulation that suc h use in

− the te r r itory of a Member Stat e in accordance with this Regulation should only be possible where and in as far as the

− Member State concer ned has decided t o expressly pro vide f or the possibility to author ise such use in its detailed r ules

− of national law . Consequently , Member Stat es remain free under this Regulation not to provide f or suc h a possibility

− at all or to only provid e f or such a possibility in respect of some of the objectives capable of justifying author ised use

− identified in this Regulation. Such national r ules should be notified to the Commission within 30 da ys of their

− adop tion.

− (38) The use of AI syste ms f or real-time remote biometr ic identification of natural persons in publicly accessible spaces

− f or the pur pose of law enf orcement necessar ily inv olves the processing of biometr ic data. The r ules of this

− Regulation that prohibit, subject t o cer tain ex ceptions, suc h use, which are based on Ar ticle 16 TFEU, should apply

− as lex specialis in respect of the r ules on the processing of biometr ic data contained in Ar ticle 10 of Directive (EU)

− 2016/680, thus regulating such use and the processing of biometr ic data inv olved in an exhaustive manner .

− Theref ore, suc h use and processing should be possible only in as f ar as it is compatible with the framework set by

− this Regulation, without there being scope, outside that framework, f or the comp etent author ities, where they act f or

− pur pose of law enf orcement, to use suc h syste ms and process suc h data in connection thereto on the grounds list ed

− in Ar ticle 10 of Directive (EU) 2016/680. In that cont ext, this Regulation is not intended to provide the leg al basis

− f or the processing of personal data under Ar ticle 8 of Directive (EU) 2016/680. Ho wever , the use of real-time remote

− biometr ic identifica tion syste ms in publicly accessible spaces f or pur poses other than law enf orcement, including by

− com petent author ities, should not be covered b y the specif ic framew ork regard ing suc h use f or the pur pose of law

− enf orcement set by this Regulation. Such use f or pur poses other than law enf orcement should theref ore not be

− subject t o the requirement of an author isation under this Regulation and the applicable detailed r ules of national law

− that ma y give effect to that author isation.

− (39) Any processing of biometr ic data and other personal data in volved in the use of AI syste ms f or biometr ic

− identification, other than in connection to the use of real-time remote biometr ic identification syste ms in publicly

− accessible spaces f or the pur pose of law enf orcement as regulated by this Regulation, should continue to comply

− with all requirements resulting from Ar ticle 10 of Directive (EU) 2016/680. For pur poses other than la w

− enf orcement, Ar ticle 9(1) of Regulation (EU) 2016/679 and Ar ticle 10(1) of Regulation (EU) 2018/1725 prohibit the

− processing of biometr ic data subject to limit ed ex ceptions as provided in those Ar ticles. In the application of Ar ticle

− 9(1) of Regulation (EU) 2016/679, the use of remot e biometr ic identifi cation f or pur poses other than la w

− enf orcement has already been subject to prohibition decisions by national data protection author ities.

− (40) In accordance with Ar ticle 6a of Protocol No 21 on the position of the Uni ted Kingdom and Ireland in respect of the

− area of freedom, secur ity and justice, as annexe d to the TEU and t o the TFEU, Ireland is not bound by the r ules laid

− do wn in Ar ticle 5(1), first subparagraph, point (g), to the exte nt it applies t o the use of biometr ic categor isation

− syste ms f or activities in the field of police cooperation and judicial cooperation in cr iminal matt ers, Ar ticle 5(1), fi rst

− subparagraph, point (d), to the exte nt it applies to the use of AI syste ms covered by that provision, Ar ticle 5(1), fi rst

− subparagraph, point (h), Ar ticle 5(2) to (6) and Ar ticle 26(10) of this Regulation adop ted on the basis of Ar ticle 16

− TFEU which relate to the processing of personal data by the Member Stat es when car r ying out activities f alling

− within the scope of Chapt er 4 or Chapt er 5 of Title V of Pa r t Three of the TFEU, where Ireland is not bound by the

− r ules gover ning the f or ms of judicial cooperation in cr iminal matt ers or police cooperation which require

− com pliance with the provisions laid do wn on the basis of Ar ticle 16 TFEU.

− (41) In accordance with Ar ticles 2 and 2a of Proto col No 22 on the position of Denmark, annexe d to the TEU and t o the

− TFEU, Denmark is not bound by r ules laid down in Ar ticle 5(1), first subparagraph, point (g), to the exte nt it applies

− t o the use of biometr ic categor isation syste ms f or activities in the fi eld of police cooperation and judicial cooperation

− in cr iminal matt ers, Ar ticle 5(1), first subparagraph, point (d), to the extent it applies t o the use of AI syste ms

− co vered b y that provision, Ar ticle 5(1), fir st subparagraph, point (h), (2) to (6) and Ar ticle 26(10) of this Regulation

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 11/144

− adop te d on the basis of Ar ticle 16 TFEU, or subject to their application, which relate to the processing of personal

− data by the Member States when car r ying out activities f alling within the scope of Chap ter 4 or Chapt er 5 of

− Title V of Par t Three of the TFEU.

− (42) In line with the presump tion of innocence, natural persons in the Uni on should alwa ys be judged on their actual

− behavi our . Natural persons should never be judg ed on AI-predicted behavio ur based solely on their prof iling,

− personality traits or char acter istics, such as nationality , place of bir th, place of residence, number of children, level of

− debt or type of car , without a reasonable suspicion of that person being in volved in a cr iminal activity based on

− objective ver ifi able f acts and without human assessment thereof. Theref ore, r isk assessments car r ied out with rega rd

− t o natural persons in order to assess the likelihood of their offending or to predict the occur rence of an actual or

− pote ntial cr iminal offence based solely on prof iling them or on assessing their personality traits and ch aracter istics

− should be prohibited. In any case, that prohibition does not refer to or tou ch upon r isk analytics that are not based

− on the profil ing of individuals or on the personality traits and charact er istics of individuals, suc h as AI systems using

− r isk analytics t o assess the like lihood of financ ial fraud by under takings on the basis of suspicious transactions or

− r isk analytic to ols to predict the likelihood of the localisation of narcotics or illicit goods by customs author ities, f or

− exam ple on the basis of known traff ic king routes.

− (43) The placing on the mark et, the putting into ser vice f or that specific pur pose, or the use of AI syste ms that create or

− expand f acial recognition databases through the untarg eted scraping of f acial images from the inter net or CCT V

− f ootage, should be prohibite d because that practice adds to the f eeling of mass sur veillance and can lead to gross

− violations of fundamental r ights, including the r ight to pr ivacy .

− (44) There are ser ious concer ns about the scientific basis of AI systems aiming to identify or infer emotions, par ticularly

− as expression of emotions var y considerably across cultures and situations, and even within a sing le individual.

− Among the ke y shor tcomings of such syste ms are the limited reliability , the lac k of specificity and the limit ed

− g eneralisability . Theref ore, AI systems identifying or inferr ing emotions or intentions of natural persons on the basis

− of their biometr ic data ma y lead t o discr iminat or y outcomes and can be intr usive to the r ights and freedoms of the

− concer ned persons. Consider ing the imbalance of power in the cont ext of w ork or education, combined with the

− intr usive nature of these syste ms, suc h systems could lead t o detr imental or unfa v ourable treatment of cer tain

− natural persons or whole groups thereof. Theref ore, the placing on the marke t, the putting into ser vice, or the use of

− AI syste ms inte nded to be used to detect the emotional state of individuals in situations related to the w orkplace and

− education should be prohibite d. That prohibition should not cover AI systems placed on the mark et str ictly f or

− medical or saf ety reasons, suc h as systems inte nded f or therapeutical use.

− (45) Practices that are prohibited by Uni on law , including data prot ection law , non-discr imination law , consumer

− prot ection law , and competition law , should not be affe cted by this Regulation.

− (46) High-r isk AI systems should only be placed on the Uni on marke t, put into ser vice or used if the y comply with

− cer tain mandato r y requirements. Those requirements should ensure that high-r isk AI systems available in the Uni on

− or whose output is other wise used in the Union do not pose unaccepta ble r isks t o im por tant Uni on public inte rests

− as recognised and protect ed by Uni on law . On the basis of the New Legislative Framework, as clar ified in the

− Commission notice ‘ The “Blue Guide” on the imp lementation of EU product r ules 2022’ (

− 20

− ), the general r ule is that

− more than one lega l act of Uni on har monisation legislation, suc h as Regulations (EU) 2017/745 (

− 21

− ) and (EU)

− 2017/746 (

− 22

− ) of the European Parliament and of the Council or Directive 2006/42/EC of the European Pa rliament

− and of the Council (

− 23

− ), ma y be applicable to one product, since the making availa ble or putting into ser vice can take

− 12/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− (

− 20

− ) OJ C 247, 29.6.2022, p. 1.

− (

− 21

− ) Regulation (EU) 2017/745 of the European Parliament and of the Council of 5 Apr il 2017 on medical devices, amending Directive

− 2001/83/EC, Regulation (EC) No 178/2002 and Regulation (EC) No 1223/2009 and repealing Council Directives 90/385/EEC and

− 93/42/EEC (OJ L 117, 5.5.2017, p. 1).

− (

− 22

− ) Regulation (EU) 2017/746 of the European Parliament and of the Council of 5 Apr il 2017 on in vitro diagnostic medical devices and

− repealing Directive 98/79/EC and Commission Decision 2010/227/EU (OJ L 117, 5.5.2017, p. 176).

− (

− 23

− ) Directive 2006/42/EC of the European Parl iament and of the Council of 17 Ma y 2006 on machiner y , and amending Directive

− 95/16/EC (OJ L 157, 9.6.2006, p. 24).

− place only when the product comp lies with all applicable Uni on har monisation legislation. T o ensure consiste ncy

− and av oid unnecessar y administrative burdens or costs, provid ers of a product that contains one or more high-r isk

− AI syste ms, to which the requirements of this Regulation and of the Union har monisation legislation listed in an

− annex t o this Regulation apply , should hav e f lexibility with regard to operational decisions on ho w t o ensure

− com pliance of a product that contains one or more AI syste ms with all applicable requirements of the Uni on

− har monisation legislation in an optima l manner . AI syste ms identified as high-r isk should be limit ed to those that

− hav e a significant har mful impact on the health, saf ety and fundamental r ights of persons in the Uni on and such

− limitation should minimise any pote ntial restr iction to inte r national trade.

− (47) AI syste ms could have an adverse impact on the health and saf ety of persons, in par ticular when suc h syste ms

− operat e as safety comp onents of products. Consiste nt with the objectives of Union har monisation legislation to

− f acilitate the free move ment of products in the inter nal market and to ensure that only saf e and other wise comp liant

− products fi nd their wa y into the market, it is imp or tant that the safety r isks that ma y be g enerated by a product as

− a whole due to its digital com ponents, including AI syste ms, are duly prevented and mitigat ed. For instance,

− increasing ly autono mous robots, whether in the cont ext of manuf actur ing or personal assistance and care should be

− able to safely operate and perfo r ms their functions in complex environments. Similarly , in the health secto r where

− the stakes f or lif e and health are par ticularly high, increasing ly sophisticat ed diagnostics systems and syste ms

− suppor ting human decisions should be reliable and accurate.

− (48) The extent of the adverse imp act caused b y the AI syste m on the fundamental r ights protect ed b y the Char ter is of

− par ticular relevance when classifying an AI syste m as high r isk. Those r ights include the r ight to human dignity ,

− respect f or pr ivate and f amily life, protection of personal data, freedom of expression and inf or mation, freedom of

− assembly and of association, the r ight to non-discr imination, the r ight to education, consumer protection, work ers’

− r ights, the r ights of persons with disabilities, gender equality , inte llectual proper ty r ights, the r ight t o an eff ective

− remedy and to a f air tr ial, the r ight of defence and the presump tion of innocence, and the r ight to good

− administration. In addition to those r ights, it is impor tant t o highlight the f act that ch ildren hav e specif ic r ights as

− enshr ined in Ar ticle 24 of the Char te r and in the Unit ed Nations Convention on the Rights of the Child, fur ther

− developed in the UNCR C General Comment No 25 as regard s the digital en vironment, both of which require

− consideration of the ch ildren’s vulnerabilities and provision of suc h protection and care as necessar y f or their

− well-being. The fundamental r ight to a high level of envir onmental protection enshr ined in the Char te r and

− im plemented in Union policies should also be considered when assessing the sever ity of the har m that an AI system

− can cause, including in relation to the health and safety of persons.

− (49) As regard s high-r isk AI systems that are safety com ponents of products or syste ms, or which are themselves

− products or systems f alling within the scope of Regulation (EC) No 300/2008 of the European Parliame nt and of the

− Council (

− 24

− ), Regulation (EU) No 167/2013 of the European Parliame nt and of the Council (

− 25

− ), Regulation

− (EU) No 168/2013 of the European Parliament and of the Council (

− 26

− ), Directive 2014/90/EU of the European

− P arliament and of the Council (

− 27

− ), Directive (EU) 2016/797 of the European Pa rliament and of the Council (

− 28

− ),

− Regulation (EU) 2018/858 of the European Pa rliament and of the Council (

− 29

− ), Regulation (EU) 2018/1139 of the

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 13/144

− (

− 24

− ) Regulation (EC) No 300/2008 of the European Parliament and of the Council of 11 March 2008 on common r ules in the f ield of

− civil aviation secur ity and repealing Regulation (EC) No 2320/2002 (OJ L 97, 9.4.2008, p. 72).

− (

− 25

− ) Regulation (EU) No 167/2013 of the European Parl iament and of the Council of 5 Febr uar y 2013 on the approv al and market

− sur veillance of agr icultural and f orestr y vehicles (OJ L 60, 2.3.2013, p. 1).

− (

− 26

− ) Regulation (EU) No 168/2013 of the European Parl iament and of the Council of 15 Januar y 2013 on the approval and market

− sur veillance of two- or three-wheel vehicles and quadr icy cles (OJ L 60, 2.3.2013, p. 52).

− (

− 27

− ) Directive 2014/90/EU of the European Parliament and of the Council of 23 July 2014 on mar ine equipment and repealing Council

− Directive 96/98/EC (OJ L 257, 28.8.2014, p. 146).

− (

− 28

− ) Directive (EU) 2016/797 of the European Parliament and of the Council of 11 Ma y 2016 on the int eroperability of the rail system

− within the European Union (OJ L 138, 26.5.2016, p. 44).

− (

− 29

− ) Regulation (EU) 2018/858 of the European Parl iament and of the Council of 30 Ma y 2018 on the approval and market sur veillance

− of motor vehicles and their trailers, and of systems, componen ts and separate tec hnical units intended f or suc h vehicles, amending

− Regulations (EC) No 715/2007 and (EC) No 595/2009 and repealing Directive 2007/46/EC (OJ L 151, 14.6.2018, p. 1).

− European P arliament and of the Council (

− 30

− ), and Regulation (EU) 2019/2144 of the European Parliament and of the

− Council (

− 31

− ), it is appropr iate to amend those acts t o ensure that the Commission tak es into account, on the basis of

− the te chnical and regulato r y specif icities of each secto r , and without interfer ing with existing gover nance, conf or mity

− assessment and enf orcement mechanisms and author ities established therein, the mandato r y requirements f or

− high-r isk AI systems laid do wn in this Regulation when adopti ng any relevant deleg ated or imp lementing acts on the

− basis of those acts.

− (50) As regards AI syste ms that are safety comp onents of products, or which are themselves products, f alling within the

− scope of cer tain Union har monisation legislation list ed in an annex to this Regulation, it is appropr iat e to classify

− them as high-r isk under this Regulation if the product concer ned undergoes the conf or mity assessment procedure

− with a third-par ty conf or mity assessment body pursuant to that relevant Uni on har monisation legislation. In

− par ticular , suc h products are machi ner y , t o ys, lif ts, equipment and prot ective systems intended f or use in pote ntially

− explosive atmospheres, radio equipment, pressure equipment, recreational craf t equipment, cablewa y installations,

− appliances bur ning g aseous fuels, medical devices, in vitro diagnostic medical devices, automot ive and aviation.

− (51) The classif ication of an AI syste m as high-r isk pursuant t o this Regulation should not necessar ily mean that the

− product whose saf ety component is the AI system, or the AI system itself as a product, is considered to be high-r isk

− under the cr iteria established in the relevant Uni on har monisation legislation that applies to the product. This is, in

− par ticular , the case f or Regulations (EU) 2017/745 and (EU) 2017/746, where a third-par ty conf or mity assessment is

− pro vided f or medium-r isk and high-r isk products.

− (52) As rega rds stand-alone AI syste ms, namely high-r isk AI syste ms other than those that are saf ety com ponents of

− products, or that are themselves products, it is appropr iat e to classify them as high-r isk if, in light of their intended

− pur pose, the y pose a high r isk of har m t o the health and safety or the fundamental r ights of persons, taking into

− account both the sever ity of the possible har m and its probability of occur rence and they are used in a number of

− specifically pre-defi ned areas specified in this Regulation. The identification of those systems is based on the same

− methodology and cr ite r ia envisag ed also f or any future amendments of the list of high-r isk AI syste ms that the

− Commission should be empo wered to adop t, via delegat ed acts, to take into account the rapid pace of te chnological

− development, as well as the potential ch ange s in the use of AI systems.

− (53) It is also impor tant to clar ify that there ma y be specific cases in which AI syste ms refer red to in pre-define d areas

− specified in this Regulation do not lead to a signifi cant r isk of har m to the lega l intere sts protect ed under those areas

− because they do not materi ally inf luence the decision-making or do not har m those interests substantially . For the

− pur poses of this Regulation, an AI system that does not mat er ially inf luence the outcome of decision-making should

− be understood to be an AI syste m that does not have an im pact on the substance, and thereby the outcome, of

− decision-making, whether human or automa ted. An AI system that does not materially inf luence the outcome of

− decision-making could include situations in which one or more of the f ollo wing conditions are fulfilled. The fir st

− suc h condition should be that the AI syste m is intende d t o perfo r m a nar row procedural task, suc h as an AI system

− that transf or ms unstr uctured data into str uctured data, an AI system that classif ies incoming documents into

− cate gor ies or an AI syste m that is used to det ect duplicate s among a large number of applications. Those tasks are of

− suc h nar row and limited nature that they pose only limited r isks which are not increased through the use of an AI

− syste m in a cont ext that is list ed as a high-r isk use in an annex to this Regulation. The second condition should be

− 14/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− (

− 30

− ) Regulation (EU) 2018/1139 of the European Parliament and of the Council of 4 July 2018 on common r ules in the f ield of civil

− av iation and establishing a European Union A viation Safety Ag ency , and amending Regulations (EC) No 2111/2005, (EC)

− No 1008/2008, (EU) No 996/2010, (EU) No 376/2014 and Directives 2014/30/EU and 2014/53/EU of the European Parl iament

− and of the Council, and repealing Regulations (EC) No 552/2004 and (EC) No 216/2008 of the European Parl iament and of the

− Council and Council Regulation (EEC) No 3922/91 (OJ L 212, 22.8.2018, p. 1).

− (

− 31

− ) Regulation (EU) 2019/2144 of the European Parliament and of the Council of 27 November 2019 on type-approval requirements

− f or motor vehicles and their trailers, and systems, components and separate tec hnical units int ended f or such vehicles, as reg ards

− their ge neral safety and the protection of vehicle occupants and vulnerable road users, amending Regulation (EU) 2018/858 of the

− European Parl iament and of the Council and repealing Regulations (EC) No 78/2009, (EC) No 79/2009 and (EC) No 661/2009 of

− the European Parl iament and of the Council and Commission Regulations (EC) No 631/2009, (EU) No 406/2010, (EU)

− No 672/2010, (EU) No 1003/2010, (EU) No 1005/2010, (EU) No 1008/2010, (EU) No 1009/2010, (EU) No 19/2011, (EU)

− No 109/2011, (EU) No 458/2011, (EU) No 65/2012, (EU) No 130/2012, (EU) No 347/2012, (EU) No 351/2012, (EU)

− No 1230/2012 and (EU) 2015/166 (OJ L 325, 16.12.2019, p. 1).

− that the task perfo r med by the AI system is intended to imp ro ve the result of a previously comp leted human activity

− that ma y be relevant f or the pur poses of the high-r isk uses list ed in an annex to this Regulation. Consider ing those

− ch aracteristics , the AI system provid es only an additional la yer to a human activity with consequently lower ed r isk .

− That condition w ould, f or exam ple, apply t o AI syste ms that are intended to impro ve the language used in previously

− draf ted documents, f or exam ple in relation to profe ssional tone, academic style of languag e or by aligning te xt to

− a cer tain brand messaging. The third condition should be that the AI syste m is intended to det ect decision-making

− patt er ns or deviations from pr ior decision-making patterns . The r isk wo uld be lower ed because the use of the AI

− syste m f ollo ws a previously complet ed human assessment which it is not meant to replace or inf luence, without

− proper human review . Such AI syste ms include f or instance those that, given a cer tain grading patt er n of a teac her ,

− can be used to ch eck ex post whether the t eacher ma y hav e deviated from the grading pattern so as to f lag potent ial

− inconsiste ncies or anomalies. The f our th condition should be that the AI syste m is intended to perf or m a task that is

− only preparat or y to an assessment relevant f or the pur poses of the AI systems listed in an annex t o this Regulation,

− thus making the possible imp act of the output of the system ver y low in t er ms of representing a r isk f or the

− assessment to f ollow . That condition covers, inte r alia, smar t solutions f or file handling, which include var ious

− functions from indexing, search ing, te xt and speech processing or linking data t o other data sources, or AI syste ms

− used f or translation of initial documents. In any case, AI systems used in high-r isk use-cases list ed in an annex to this

− Regulation should be considered to pose signifi cant r isks of har m to the health, safety or fundamental r ights if the AI

− syste m imp lies profiling within the meaning of Ar ticle 4, point (4) of Regulation (EU) 2016/679 or Ar ticle 3,

− point (4) of Directive (EU) 2016/680 or Ar ticle 3, point (5) of Regulation (EU) 2018/1725. T o ensure traceability

− and transparency , a provid er who considers that an AI syste m is not high-r isk on the basis of the conditions refe r red

− t o above should dra w up documentation of the assessment bef ore that syste m is placed on the marke t or put into

− ser vice and should provide that documentation to national compet ent author ities upon request. Such a provid er

− should be obliged to register the AI syste m in the EU database established under this Regulation. With a view to

− pro viding fur ther guidance f or the practical imp lementation of the conditions under which the AI syste ms listed in

− an annex to this Regulation are, on an excep tional basis, non-high-r isk, the Commission should, af te r consulting the

− Board, provid e guidelines specifying that practical implementation, comp let ed b y a compre hensive list of practical

− exam ples of use cases of AI syste ms that are high-r isk and use cases that are not.

− (54) As biometr ic data constitute s a special categor y of personal data, it is appropr iat e to classify as high-r isk several

− cr itical-use cases of biometr ic syste ms, insofar as their use is per mitte d under relevant Uni on and national law .

− T ec hnical inaccuracies of AI systems intende d f or the remote biometr ic identifica tion of natural persons can lead to

− biased results and entail discr iminat or y eff ects. The r isk of suc h biased results and discr iminat or y effects is

− par ticularly relevant with regard to age , ethnicity , race, sex or disabilities. Remote biometr ic identifica tion syste ms

− should theref ore be classif ied as high-r isk in view of the r isks that they pose. Such a classif ication excludes AI

− syste ms inte nded to be used f or biometr ic ver ifi cation, including authentication, the sole pur pose of which is to

− confir m that a specif ic natural person is who that person claims t o be and to confir m the identity of a natural person

− f or the sole pur pose of hav ing access to a ser vice, unlocking a device or having secure access to premises. In addition,

− AI syste ms intended to be used f or biometr ic categor isation according to sensitive attr ibut es or ch aracter istics

− prot ected under Ar ticle 9(1) of Regulation (EU) 2016/679 on the basis of biometr ic data, in so far as these are not

− prohibite d under this Regulation, and emotion recognition systems that are not prohibite d under this Regulation,

− should be classif ied as high-r isk. Biometr ic syste ms which are intende d t o be used solely f or the pur pose of enabling

− cybersecur ity and personal data prot ection measures should not be considered to be high-r isk AI syste ms.

− (55) As regards the management and operation of cr itical infrastr ucture, it is appropr iate to classify as high-r isk the AI

− syste ms inte nded to be used as safety components in the managem ent and operation of cr itical digital infrastr ucture

− as list ed in point (8) of the Annex to Directive (EU) 2022/2557, road traffic and the supply of wate r , g as, heating and

− electr icity , since their f ailure or malfunctioning ma y put at r isk the life and health of persons at larg e scale and lead t o

− appreciable disr up tions in the ordinar y conduct of social and economic activities. Saf ety compo nents of cr itical

− infrastr ucture, including cr itical digital infrastr ucture, are systems used to directly protect the physica l inte gr ity of

− cr itical infrastr ucture or the health and safety of persons and proper ty but which are not necessar y in order f or the

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 15/144

− syste m to function. The f ailure or malfunctioning of suc h comp onents might directly lead to r isks to the phys ical

− inte gr ity of cr itical infrastr ucture and thus t o r isks to health and safety of persons and proper ty . Compo nents

− inte nded to be used solely f or cybersecur ity pur poses should not qualify as safety compo nents. Examples of saf ety

− com ponents of suc h cr itical infrastr ucture ma y include syste ms f or monitoring wate r pressure or fire alar m

− controlling systems in cloud comp uting centres.

− (56) The deplo yment of AI systems in education is impor tant to promot e high-quality digital education and training and

− t o allo w all lear ners and te ache rs to acquire and share the necessar y digital skills and comp etences, including media

− lite racy , and cr itical thinking, to take an active par t in the economy , society , and in democratic processes. How ever ,

− AI systems used in education or v ocational training, in par ticular f or det er mining access or admission, f or assigning

− persons to educational and v ocational training institutions or programmes at all levels, f or evaluating lear ning

− outcomes of persons, f or assessing the appropr iate level of education f or an individual and mat er ially inf luencing the

− level of education and training that individuals will receive or will be able to access or f or monitoring and detecting

− prohibite d behaviour of students dur ing tests should be classif ied as high-r isk AI systems, since the y ma y determi ne

− the educational and profe ssional course of a person’s life and theref ore ma y affe ct that person’s ability to secure

− a livelihood. When improperly designed and used, suc h syste ms ma y be par ticularly intr usive and ma y violate the

− r ight to education and training as well as the r ight not to be discr iminated aga inst and per petuate histor ical patt er ns

− of discr imination, f or exam ple ag ainst women, cer tain age groups, persons with disabilities, or persons of cer tain

− racial or ethnic or igins or sexual or ientation.

− (57) AI systems used in emplo yment, wo rkers management and access to self-emplo yment, in par ticular f or the

− recr uitment and selection of persons, f or making decisions affe cting te r ms of the work -related relationship,

− promotion and term ination of w ork -relate d contractual relationships, f or allocating tasks on the basis of individual

− behavi our , personal traits or ch aracter istics and f or monitoring or evaluation of persons in work -relate d contractual

− relationships, should also be classif ied as high-r isk, since those systems ma y hav e an appreciable im pact on future

− career prospects, livelihoods of those persons and work ers’ r ights. Relevant wo rk -related contractual relationships

− should, in a meaningful manner , in volve emp lo y ees and persons provid ing ser vices through platf or ms as refe r red to

− in the Commission W ork Programme 2021. Throughout the recr uitment process and in the evaluation, promotion,

− or retention of persons in w ork -related contractual relationships, suc h syste ms ma y per petuat e historical patterns of

− discr imination, f or exam ple against w omen, cer tain age groups, persons with disabilities, or persons of cer tain racial

− or ethnic or igins or sexual or ientation. AI syste ms used to monitor the perf or mance and behaviour of suc h persons

− ma y also under mine their fundamental r ights to data protection and pr ivacy .

− (58) Another area in which the use of AI syste ms deser ves special consideration is the access to and enjo yment of cer tain

− essential pr ivate and public ser vices and benef its necessar y f or people to fully par ticipate in society or to imp rove

− one’s standard of living. In par ticular , natural persons applying f or or receiving essential public assistance benef its

− and ser vices from public author ities namely healthcare ser vices, social secur ity benef its, social ser vices providing

− prot ection in cases suc h as maternity , illness, industr ial accidents, dependency or old age and loss of emp lo yment

− and social and housing assistance, are typically dependent on those benef its and ser vices and in a vulnerable position

− in relation to the responsible author ities. If AI syste ms are used f or det er mining whether suc h benefits and ser vices

− should be granted, denied, reduced, revok ed or reclaimed by author ities, including whether beneficiar ies are

− legitimat ely entitled to suc h benefi ts or ser vices, those syste ms ma y hav e a significant impact on persons’ livelihood

− and ma y infr inge their fundamental r ights, suc h as the r ight t o social protection, non-discr imination, human dignity

− or an effe ctive remedy and should theref ore be classif ied as high-r isk. Nonetheless, this Regulation should not

− ham per the development and use of innovative approac hes in the public administration, which w ould stand to

− benefit from a wider use of compliant and safe AI systems, provid ed that those syste ms do not entail a high r isk to

− leg al and natural persons. In addition, AI syste ms used to evaluate the credit score or creditwor thiness of natural

− persons should be classif ied as high-r isk AI systems, since the y determi ne those persons’ access to financ ial resources

− or essential ser vices suc h as housing, electr icity , and te lecommunication ser vices. AI syste ms used f or those pur poses

− ma y lead t o discr imination between persons or groups and ma y per petuat e histor ical patt er ns of discr imination,

− suc h as that based on racial or ethnic or igins, g ender , disabilities, age or sexual or ientation, or ma y create new f or ms

− of discr iminatory im pacts. However , AI systems provid ed f or by Uni on law f or the pur pose of det ecting fraud in the

− offe r ing of financ ial ser vices and f or pr udential pur poses to calculate credit institutions’ and insurance under takings ’

− capital requirements should not be considered t o be high-r isk under this Regulation. Moreove r , AI systems intended

− 16/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− t o be used f or r isk assessment and pr icing in relation to natural persons f or health and life insurance can also have

− a significant impact on persons’ livelihood and if not duly designed, developed and used, can infr inge their

− fundamental r ights and can lead t o ser ious consequences f or people ’s lif e and health, including financ ial exclusion

− and discr imination. Finally , AI systems used to evaluate and classify emerge ncy calls by natural persons or to

− dispatc h or establish pr ior ity in the dispatch ing of emerg ency fir st response ser vices, including by police, fir efighters

− and medical aid, as well as of emerge ncy healthcare patient tr iage syste ms, should also be classif ied as high-r isk since

− the y make decisions in ver y cr itical situations f or the lif e and health of persons and their proper ty .

− (59) Given their role and responsibility , actions b y la w enf orcement author ities inv olving cer tain uses of AI systems are

− ch aracterised b y a signifi cant degree of power imbalance and ma y lead to sur veillance, ar rest or depr ivation of

− a natural person’s liber ty as well as other adverse im pacts on fundamental r ights guaranteed in the Char te r . In

− par ticular , if the AI syste m is not trained with high-quality data, does not meet adequate requirements in ter ms of its

− perf or mance, its accuracy or robustness, or is not properly designed and te sted bef ore being put on the market or

− other wise put into ser vice, it ma y sing le out people in a discr iminat or y or other wise incor rect or unjust manner .

− Fur ther more, the ex ercise of imp or tant procedural fundamental r ights, suc h as the r ight to an eff ective remedy and

− t o a f air tr ial as well as the r ight of defence and the presump tion of innocence, could be hampered, in par ticular ,

− where such AI syste ms are not suffi ciently transparent, explainable and documente d. It is theref ore appropr iate to

− classify as high-r isk, insofa r as their use is per mitte d under relevant Union and national law , a number of AI syste ms

− inte nded t o be used in the law enf orcement cont ext where accuracy , reliability and transparency is par ticularly

− im por tant to av oid adverse im pacts, retain public tr ust and ensure accountability and eff ective redress. In view of the

− nature of the activities and the r isks relating thereto , those high-r isk AI systems should include in par ticular AI

− syste ms inte nded t o be used by or on behalf of law enf orcement author ities or by Union institutions, bodies, offices,

− or age ncies in suppor t of law enf orcement author ities f or assessing the r isk of a natural person t o become a victim of

− cr iminal offenc es, as polygraphs and similar too ls, f or the evaluation of the reliability of evidence in in the course of

− in vestig ation or prosecution of cr iminal offences, and, insofar as not prohibited under this Regulation, f or assessing

− the r isk of a natural person offe nding or reoffe nding not solely on the basis of the prof iling of natural persons or the

− assessment of personality traits and ch aracter istics or the past cr iminal behavio ur of natural persons or groups, f or

− profiling in the course of detection, investig ation or prosecution of cr iminal offences. AI syste ms specifically

− inte nded to be used f or administrative proceedings b y tax and custo ms author ities as well as by fin ancial intellig ence

− units car r ying out administrative tasks analysing inf or mation pursuant t o Uni on anti-money launder ing law should

− not be classif ied as high-r isk AI syste ms used by law enf orcement author ities f or the pur pose of prevention,

− det ection, in vestig ation and prosecution of cr iminal offenc es. The use of AI too ls by la w enf orcement and other

− relevant author ities should not become a f actor of inequality , or exclus ion. The imp act of the use of AI to ols on the

− defe nce r ights of suspects should not be ignored, in par ticular the diff iculty in obtaining meaningful inf or mation on

− the functioning of those systems and the resulting difficulty in ch allenging their results in cour t, in par ticular by

− natural persons under investig ation.

− (60) AI systems used in migration, asylum and border control managem ent affe ct persons who are of t en in par ticularly

− vulnerable position and who are dependent on the outcome of the actions of the comp etent public author ities. The

− accuracy , non-discr iminat or y nature and transparency of the AI systems used in those contexts are theref ore

− par ticularly im por tant to guarante e respect f or the fundamental r ights of the aff ected persons, in par ticular their

− r ights t o free move ment, non-discr imination, prot ection of pr ivate life and personal data, international prot ection

− and good administration. It is theref ore appropr iate to classify as high-r isk, insofa r as their use is per mitted under

− relevant Uni on and national law , AI syste ms intended to be used by or on behalf of compet ent public author ities or

− b y Uni on institutions, bodies, off ices or age ncies ch arged with tasks in the fi elds of migration, asylum and border

− control management as polygraphs and similar t ools, f or assessing cer tain r isk s posed b y natural persons ent er ing

− the terr itory of a Member Stat e or applying f or visa or asylum, f or assisting compet ent public author ities f or the

− examination, including related assessment of the reliability of evidence, of applications f or asylum, visa and residence

− per mits and associat ed comp laints with regard to the objective to establish the eligibility of the natural persons

− applying f or a status, f or the pur pose of detecting, recognising or identifying natural persons in the context of

− migration, asylum and border control manag ement, with the excep tion of ver ification of tra vel documents. AI

− syste ms in the area of migration, asylum and border control management covered by this Regulation should comply

− with the relevant procedural requirements set by the Regulation (EC) No 810/2009 of the European Pa rliament and

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 17/144

− of the Council (

− 32

− ), the Directive 2013/32/EU of the European Parliament and of the Council (

− 33

− ), and other relevant

− Uni on law . The use of AI systems in migration, asylum and border control management should, in no circumstances,

− be used by Member Stat es or Uni on institutions, bodies, off ices or age ncies as a means to circum vent their

− inte r national obligations under the UN Conve ntion relating to the Status of Refuge es done at Geneva on 28 July

− 1951 as amended by the Prot ocol of 31 Januar y 1967. Nor should they be used to in any wa y infr ing e on the

− pr inciple of non-ref oulement, or t o deny safe and effective legal avenues into the ter r itory of the Union, including

− the r ight t o inter national prot ection.

− (61) Cer tain AI systems intended f or the administration of justice and democratic processes should be classified as

− high-r isk , consider ing their potent ially signifi cant im pact on democracy , the r ule of law , individual freedoms as well

− as the r ight to an effe ctive remedy and t o a f air tr ial. In par ticular , to address the r isks of potential biases, er rors and

− opacity , it is appropr iate t o qualify as high-r isk AI syste ms inte nded to be used by a judicial author ity or on its behalf

− t o assist judicial author ities in researching and inte r preting f acts and the law and in applying the law to a concrete set

− of f acts. AI syste ms inte nded to be used b y alt er native dispute resolution bodies f or those pur poses should also be

− considered to be high-r isk when the outcomes of the alter native dispute resolution proceedings produce leg al effe cts

− f or the par ties. The use of AI tools can suppor t the decision-making po wer of judges or judicial independence, but

− should not replace it: the final decision-making must remain a human-dr iven activity . The classification of AI

− syste ms as high-r isk should not, ho wever , extend to AI syste ms inte nded f or purely ancillar y administrative activities

− that do not affect the actual administration of justice in individual cases, suc h as anonymisa tion or

− pseudon ymisation of judicial decisions, documents or data, communication between personnel, administrative tasks.

− (62) Without prejudice to the r ules pro vided f or in Regulation (EU) 2024/900 of the European Pa rliament and of the

− Council (

− 34

− ), and in order to address the r isk s of undue exte r nal interfere nce with the r ight to v ote enshr ined in

− Ar ticle 39 of the Char te r , and of adverse effe cts on democracy and the r ule of law , AI syste ms intended to be used to

− inf luence the outcome of an election or refe rendum or the v oting behaviour of natural persons in the exercise of

− their v ote in elections or referenda should be classif ied as high-r isk AI systems with the ex ception of AI syste ms

− whose output natural persons are not directly exposed to, suc h as tools used to organise, optimise and str ucture

− political cam paigns from an administrative and logistical point of view .

− (63) The fact that an AI syste m is classif ied as a high-r isk AI system under this Regulation should not be inter preted as

− indicating that the use of the system is law ful under other acts of Uni on law or under national law compatible with

− Uni on law , such as on the prot ection of personal data, on the use of polygraphs and similar t ools or other systems to

− det ect the emotional stat e of natural persons. Any suc h use should continue to occur solely in accordance with the

− applicable requirements resulting from the Char t er and from the applicable acts of secondar y Uni on law and

− national law . This Regulation should not be understood as provid ing f or the lega l ground f or processing of personal

− data, including special categor ies of personal data, where relevant, unless it is specifically other wise provid ed f or in

− this Regulation.

− (64) T o mitigat e the r isks from high-r isk AI syste ms placed on the marke t or put into ser vice and t o ensure a high level of

− tr ustwo r thiness, cer tain mandat or y requirements should apply to high-r isk AI systems, taking into account the

− inte nded pur pose and the cont ext of use of the AI syste m and according to the r isk -management system to be

− established by the pro vider . The measures adop ted b y the providers to comply with the mandato r y requirements of

− this Regulation should take into account the g enerally ackno wledged state of the ar t on AI, be propor tionate and

− eff ective t o meet the objectives of this Regulation. Based on the New Legislative Framework, as clar ified in

− Commission notice ‘ The “Blue Guide” on the im plementation of EU product r ules 2022’, the g eneral r ule is that

− more than one legal act of Union har monisation legislation ma y be applicable to one product, since the making

− a vailable or putting into ser vice can take place only when the product complie s with all applicable Uni on

− har monisation legislation. The hazards of AI systems covered by the requirements of this Regulation concer n

− diff erent aspects than the existing Union har monisation legislation and theref ore the requirements of this Regulation

− w ould complement the existing body of the Uni on har monisation legislation. For exam ple, machi ner y or medical

− devices products incor porating an AI system might present r isks not addressed by the essential health and safety

− 18/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− (

− 32

− ) Regulation (EC) No 810/2009 of the European Parliament and of the Council of 13 July 2009 establishing a Community Code on

− Visas (Visa Code) (OJ L 243, 15.9.2009, p. 1).

− (

− 33

− ) Directive 2013/32/EU of the European Parl iament and of the Council of 26 June 2013 on common procedures f or granting and

− withdraw ing int er national prot ection (OJ L 180, 29.6.2013, p. 60).

− (

− 34

− ) Regulation (EU) 2024/900 of the European parliament and of the Council of 13 March 2024 on the transparency and targe ting of

− political adver tising (OJ L, 2024/900, 20.3.2024, ELI: http://data.europa.eu/eli/reg/2024/900/oj).

− requirements set out in the relevant Union har monised legislation, as that sectoral la w does not deal with r isks

− specific to AI systems. This calls f or a simultaneous and comp lementar y application of the var ious legislative acts. T o

− ensure consist ency and to av oid an unnecessar y administrative burden and unnecessar y costs, provid ers of a product

− that contains one or more high-r isk AI system, to which the requirements of this Regulation and of the Uni on

− har monisation legislation based on the New Legislative Framewo rk and listed in an annex t o this Regulation apply ,

− should hav e f lexibility with regard to operational decisions on how to ensure compliance of a product that contains

− one or more AI syste ms with all the applicable requirements of that Uni on har monised legislation in an op timal

− manner . That f lexibility could mean, f or exam ple a decision by the pro vider to inte grate a par t of the necessar y

− t esting and repor ting processes, inf or mation and documentation required under this Regulation into already existing

− documentation and procedures required under existing Uni on har monisation legislation based on the New

− Legislative Framework and listed in an annex to this Regulation. This should not, in any wa y , under mine the

− oblig ation of the provid er to comply with all the applicable requirements.

− (65) The r isk -managem ent syste m should consist of a continuous, ite rative process that is planned and r un throughout

− the entire lif ecy cle of a high-r isk AI syste m. That process should be aimed at identifying and mitigating the relevant

− r isks of AI syste ms on health, saf ety and fundamental r ights. The r isk -manag ement system should be regularly

− reviewed and update d to ensure its continuing effe ctiveness, as well as justifi cation and documentation of any

− significant decisions and actions take n subject to this Regulation. This process should ensure that the provid er

− identifies r isks or adverse im pacts and imp lements mitig ation measures f or the kno wn and reasonably f oreseeable

− r isks of AI syste ms to the health, safety and fundamental r ights in light of their intended pur pose and reasonably

− f oreseeable misuse, including the possible r isks ar ising from the inte raction between the AI system and the

− en vironment within which it operates. The r isk -management syste m should adop t the most appropr iate

− r isk -manag ement measures in light of the state of the ar t in AI. When identifying the most appropr iate

− r isk -manag ement measures, the provid er should document and explain the choices made and, when relevant,

− in v olve exper ts and exter nal stak eholders. In identifying the reasonably f oreseeable misuse of high-r isk AI systems,

− the provid er should cover uses of AI syste ms which, while not directly co vered by the intende d pur pose and

− pro vided f or in the instr uction f or use ma y never theless be reasonably expected t o result from readily predictable

− human behavi our in the cont ext of the specif ic charact er istics and use of a par ticular AI syste m. Any kno wn or

− f oreseeable circumstances related to the use of the high-r isk AI syste m in accordance with its inte nded pur pose or

− under conditions of reasonably f oreseeable misuse, which ma y lead t o r isks to the health and safety or fundamental

− r ights should be included in the instr uctions f or use that are provid ed by the provider . This is to ensure that the

− deplo y er is aw are and tak es them into account when using the high-r isk AI syste m. Identifying and im plementing

− r isk mitigation measures f or f oreseeable misuse under this Regulation should not require specific additional training

− f or the high-r isk AI syste m by the provid er to address f oreseeable misuse. The provider s ho wever are encourage d to

− consider suc h additional training measures t o mitiga te reasonable f oreseeable misuses as necessar y and appropr iate.

− (66) Requirements should apply to high-r isk AI syste ms as regard s r isk management, the quality and relevance of data

− sets used, te chnical documentation and record-k eeping, transparency and the provision of inf or mation to deplo y ers,

− human ove rsight, and robustness, accuracy and cybersecur ity . Those requirements are necessar y to eff ectively

− mitig ate the r isks f or health, safety and fundamental r ights. As no other less trade restr ictive measures are reasonably

− a vailable those requirements are not unjustified restr ictions to trade.

− (67) High-quality data and access to high-quality data pla ys a vital role in providing str ucture and in ensur ing the

− perf or mance of many AI systems, especially when tec hniques inv olving the training of models are used, with a view

− t o ensure that the high-r isk AI syste m per f or ms as inte nded and safely and it does not become a source of

− discr imination prohibite d by Uni on law . High-quality data sets f or training, validation and te sting require the

− im plementation of appropr iat e data gover nance and managem ent practices. Data sets f or training, validation and

− t esting, including the labels, should be relevant, sufficiently representative, and t o the best extent possible free of

− er rors and complet e in view of the intended pur pose of the system. In order to f acilitate comp liance with Uni on data

− prot ection law , suc h as Regulation (EU) 2016/679, data gove r nance and management practices should include, in

− the case of personal data, transparency about the or iginal pur pose of the data collection. The data sets should also

− hav e the appropr iat e statistical proper ties, including as regard s the persons or groups of persons in relation to whom

− the high-r isk AI syste m is inte nded to be used, with specific atte ntion to the mitigation of possible biases in the data

− sets, that are like ly t o affect the health and saf ety of persons, hav e a negative imp act on fundamental r ights or lead to

− discr imination prohibite d under Union law , especially where data outputs inf luence in puts f or future operations

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 19/144

− (f eedbac k loops). Biases can f or example be inherent in underlying data sets, especially when histor ical data is being

− used, or g enerated when the syste ms are implement ed in real world settings. Results provided b y AI systems could be

− inf luenced by such inherent biases that are inclined t o gradually increase and thereby per petuate and amplify existing

− discr imination, in par ticular f or persons belonging to cer tain vulnerable groups, including racial or ethnic groups.

− The requirement f or the data sets t o be to the best exte nt possible complet e and free of er rors should not affect the

− use of pr ivacy-preser ving te chniques in the cont ext of the development and te sting of AI systems. In par ticular , data

− sets should tak e into account, to the extent required by their intended pur pose, the f eatures, ch aracter istics or

− elements that are par ticular to the specif ic g eographical, contextual, behavi oural or functional setting which the AI

− syste m is inte nded to be used. The requirements related to data gove r nance can be complied with b y hav ing recourse

− t o third par ties that offe r cer tif ied compliance ser vices including ver ifi cation of data gover nance, data set inte gr ity ,

− and data training, validation and te sting practices, as f ar as comp liance with the data requirements of this Regulation

− are ensured.

− (68) For the development and assessment of high-r isk AI systems, cer tain actors, such as provid ers, notif ied bodies and

− other relevant entities, such as European Digital Innovation Hubs, te sting exper imentation f acilities and researchers ,

− should be able to access and use high-quality data sets within the fie lds of activities of those actor s which are related

− t o this Regulation. European common data spaces established by the Commission and the f acilitation of data shar ing

− between businesses and with gover nment in the public intere st will be instr umental to provide tr ustful, accountable

− and non-discr iminat or y access to high-quality data f or the training, validation and testing of AI syste ms. For

− exam ple, in health, the European health data space will facilitat e non-discr iminatory access to health data and the

− training of AI algor ithms on those data sets, in a pr ivacy-preser ving, secure, timely , transparent and tr ustw or th y

− manner , and with an appropr iate institutional gove r nance. Relevant comp etent author ities, including sect oral ones,

− pro viding or suppor ting the access to data ma y also suppor t the pro vision of high-quality data f or the training,

− validation and testing of AI syste ms.

− (69) The r ight to pr ivacy and to protection of personal data must be guarante ed throughout the entire lif ecy cle of the AI

− syste m. In this regard , the pr inciples of data minimisation and data protection b y design and by default, as set out in

− Uni on data protect ion law , are applicable when personal data are processed. Measures take n by provid ers to ensure

− com pliance with those pr inciples ma y include not only anonymisati on and encr ypt ion, but also the use of

− t echnology that per mits algor ithms to be brought to the data and allows training of AI systems without the

− transmission between par ties or cop ying of the raw or str uctured data themselves, without prejudice to the

− requirements on data gover nance provid ed f or in this Regulation.

− (70) In order t o protect the r ight of others from the discr imination that might result from the bias in AI syste ms, the

− pro viders should, ex ceptiona lly , to the extent that it is str ictly necessar y f or the pur pose of ensur ing bias detection

− and cor rection in relation to the high-r isk AI systems, subject to appropr iate safeguards f or the fundamental r ights

− and freedoms of natural persons and f ollowi ng the application of all applicable conditions laid do wn under this

− Regulation in addition to the conditions laid down in Regulations (EU) 2016/679 and (EU) 2018/1725 and Directive

− (EU) 2016/680, be able to process also special cate gor ies of personal data, as a matt er of substantial public interest

− within the meaning of Ar ticle 9(2), point (g) of Regulation (EU) 2016/679 and Ar ticle 10(2), point (g) of Regulation

− (EU) 2018/1725.

− (71) Ha ving comp rehensible inf or mation on ho w high-r isk AI syste ms hav e been developed and how they perf or m

− throughout their lifetime is essential to enable traceability of those syste ms, ver ify compliance with the requirements

− under this Regulation, as well as monitoring of their operations and post mark et monitoring. This requires keeping

− records and the availability of te chnical documentation, containing inf or mation which is necessar y to assess the

− com pliance of the AI syste m with the relevant requirements and f acilitate post market monitorin g. Such inf or mation

− should include the general charact er istics, capabilities and limitations of the system, algor ithms, data, training,

− t esting and validation processes used as well as documentation on the relevant r isk -manag ement system and drawn

− in a clear and comp rehensive f or m. The tec hnical documentation should be k ept up t o date, appropr iately

− throughout the lif etime of the AI syste m. Fur ther more, high-r isk AI systems should te ch nically allo w f or the

− auto matic recording of events, b y means of logs, over the duration of the lif etime of the syste m.

− 20/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− (72) T o address concer ns relate d to opacity and complexity of cer tain AI systems and help deplo yers to fulfil their

− oblig ations under this Regulation, transparency should be required f or high-r isk AI syste ms bef ore they are placed

− on the market or put it into ser vice. High-r isk AI systems should be designed in a manner t o enable deplo y ers to

− understand ho w the AI syste m works, evaluate its functionality , and comp rehend its strengths and limitations.

− High-r isk AI systems should be accom panied by appropr iate inf or mation in the f or m of instr uctions of use. Such

− inf or mation should include the charact er istics, capabilities and limitations of perfo r mance of the AI system. Those

− w ould cover inf or mation on possible kno wn and f oreseeable circumstances related to the use of the high-r isk AI

− syste m, including deplo y er action that ma y inf luence syste m behavio ur and perfor mance, under which the AI system

− can lead to r isks to health, saf ety , and fundamental r ights, on the ch anges that hav e been pre-deter mined and

− assessed f or conf or mity by the pro vider and on the relevant human oversight measures, including the measures to

− f acilitate the inter pretation of the outputs of the AI syste m by the deplo y ers. T ransparency , including the

− accom pan ying instr uctions f or use, should assist deplo y ers in the use of the syste m and suppor t inf or med decision

− making b y them. Deplo y ers should, inter alia, be in a better position to mak e the cor rect ch oice of the syste m that

− the y inte nd to use in light of the obligations applicable to them, be educat ed about the inte nded and precluded uses,

− and use the AI system cor rectly and as appropr iate. In order to enhance legibility and accessibility of the inf or mation

− included in the instr uctions of use, where appropr iat e, illustrative exam ples, f or instance on the limitations and on

− the intende d and precluded uses of the AI syste m, should be included. Provi ders should ensure that all

− documentation, including the instr uctions f or use, contains meaningful, compre hensive, accessible and

− understandable inf or mation, taking into account the needs and f oreseeable kno wledge of the targe t deplo y ers.

− Instr uctions f or use should be made a vailable in a language which can be easily understood by targ et deplo y ers, as

− det er mined by the Member Stat e concer ned.

− (73) High-r isk AI systems should be designed and developed in such a wa y that natural persons can ove rsee their

− functioning, ensure that they are used as intended and that their imp acts are addressed over the system’s lifecy cle. T o

− that end, appropr iate human oversight measures should be identifie d by the pro vider of the syste m bef ore its placing

− on the market or putting into ser vice. In par ticular , where appropr iate, suc h measures should guarantee that the

− syste m is subject to in-built operational constraints that cannot be overr idden by the system itself and is responsive

− t o the human operat or , and that the natural persons to whom human oversight has been assigned have the necessar y

− com petence, training and author ity to car r y out that role. It is also essential, as appropr iate, t o ensure that high-r isk

− AI syste ms include mechanisms to guide and inf or m a natural person to whom human oversight has been assigned

− t o mak e inf or med decisions if, when and ho w to inte r vene in order to av oid nega tive consequences or r isks , or stop

− the syste m if it does not per f or m as intended. Consider ing the significant consequences f or persons in the case of an

− incor rect matc h by cer tain biometr ic identification systems, it is appropr iat e to provide f or an enhanced human

− o versight requirement f or those systems so that no action or decision ma y be take n by the deplo yer on the basis of

− the identifica tion resulting from the system unless this has been separate ly ver ified and conf ir med by at least tw o

− natural persons. Those persons could be from one or more entities and include the person operating or using the

− syste m. This requirement should not pose unnecessar y burden or dela ys and it could be suffi cient that the separate

− ver ificati ons by the diffe rent persons are auto matically recorded in the logs g enerated by the syste m. Given the

− specificities of the areas of law enf orcement, migration, border control and asylum, this requirement should not

− apply where Uni on or national law considers the application of that requirement to be dispropor tionate.

− (74) High-r isk AI syste ms should perf or m consiste ntly throughout their lif ecy cle and meet an appropr iate level of

− accuracy , robustness and cybersecur ity , in light of their inte nded pur pose and in accordance with the g enerally

− ac knowledg ed stat e of the ar t. The Commission and relevant organisations and stak eholders are encourage d to take

− due consideration of the mitiga tion of r isks and the nega tive impacts of the AI system. The expecte d level of

− perf or mance metr ics should be declared in the accom panying instr uctions of use. Providers are urg ed to

− communicate that inf or mation t o deplo yers in a clear and easily understandable wa y , free of misunderstandings or

− misleading statements. Union law on lega l metrology , including Directives 2014/31/EU (

− 35

− ) and 2014/32/EU (

− 36

− ) of

− the European Parliament and of the Council, aims to ensure the accuracy of measurements and to help the

− transparency and fairne ss of commercial transactions. In that context, in cooperation with relevant stakeh olders and

− org anisation, such as metrology and bench marking author ities, the Commission should encourage , as appropr iate,

− the development of bench mark s and measurement methodologies f or AI systems. In doing so, the Commission

− should take note and collaborate with inter national par tners wo rking on metrology and relevant measurement

− indicato rs relating to AI.

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 21/144

− (

− 35

− ) Directive 2014/31/EU of the European Parliament and of the Council of 26 Febr uar y 2014 on the har monisation of the la ws of the

− Member State s relating to the making available on the market of non-aut omatic weighing instr uments (OJ L 96, 29.3.2014, p. 107).

− (

− 36

− ) Directive 2014/32/EU of the European Parliament and of the Council of 26 Febr uar y 2014 on the har monisation of the la ws of the

− Member States relating t o the making available on the market of measur ing instr uments (OJ L 96, 29.3.2014, p. 149).

− (75) T ec hnical robustness is a ke y requirement f or high-r isk AI systems. They should be resilient in relation to har mful or

− other wise undesirable behavi our that ma y result from limitations within the syste ms or the envir onment in which

− the systems operate (e.g. er rors, faults, inconsiste ncies, unexpected situations). Theref ore, tec hnical and

− org anisational measures should be take n to ensure robustness of high-r isk AI systems, f or exam ple b y designing

− and developing appropr iate te chnical solutions to prevent or minimise har mful or other wise undesirable behavi our .

− Those te chnical solution ma y include f or instance mec hanisms enabling the system to saf ely inter r up t its operation

− (f ail-saf e plans) in the presence of cer tain anomalies or when operation take s place outside cer tain predeter mined

− boundar ies. Fa ilure to protect aga inst these r isks could lead to safety im pacts or nega tively aff ect the fundamental

− r ights, f or exam ple due to er roneous decisions or wrong or biased outputs g enerated by the AI syste m.

− (76) Cybersecur ity pla ys a cr ucial role in ensur ing that AI syste ms are resilient aga inst attem pts to alt er their use,

− behavi our , perf or mance or comp romise their secur ity proper ties by malicious third par ties exploiting the syste m’s

− vulnerabilities. Cyberattacks ag ainst AI syste ms can leverage AI specific assets, suc h as training data sets (e.g. data

− poisoning) or trained models (e.g. adversar ial attacks or membership inf erence), or exploit vulnerabilities in the AI

− syste m’s digital assets or the underlying ICT infrastr ucture. T o ensure a level of cybersecur ity appropr iate t o the r isks ,

− suitable measures, suc h as secur ity controls, should theref ore be take n by the provid ers of high-r isk AI systems, also

− taking into account as appropr iate the underlying ICT infrastr ucture.

− (77) Without prejudice to the requirements related to robustness and accuracy set out in this Regulation, high-r isk AI

− syste ms which f all within the scope of a regulation of the European Parliament and of the Council on hor izontal

− cybersecur ity requirements f or products with digital elements, in accordance with that regulation ma y demonstrate

− com pliance with the cybersecur ity requirements of this Regulation by fulf illing the essential cybersecur ity

− requirements set out in that regulation. When high-r isk AI systems fulf il the essential requirements of a regulation of

− the European Parliament and of the Council on hor izontal cybersecur ity requirements f or products with digital

− elements, the y should be deemed compliant with the cybersecur ity requirements set out in this Regulation in so f ar

− as the ac hievement of those requirements is demonstrated in the EU declaration of conf or mity or par ts thereof

− issued under that regulation. T o that end, the assessment of the cybersecur ity r isks , associated to a product with

− digital elements classif ied as high-r isk AI syste m according to this Regulation, car r ied out under a regulation of the

− European Parliament and of the Council on hor izontal cybersecur ity requirements f or products with digital

− elements, should consider r isks to the cyber resilience of an AI system as regard s attem pts by unauthor ised third

− par ties to alt er its use, behavio ur or perf or mance, including AI specif ic vulnerabilities suc h as data poisoning or

− adversar ial attacks, as well as, as relevant, r isks to fundamental r ights as required b y this Regulation.

− (78) The conf or mity assessment procedure provid ed by this Regulation should apply in relation t o the essential

− cybersecur ity requirements of a product with digital elements covered by a regulation of the European Pa rliament

− and of the Council on hor izontal cybersecur ity requirements f or products with digital elements and classified as

− a high-r isk AI system under this Regulation. However , this r ule should not result in reducing the necessar y level of

− assurance f or cr itical products with digital elements covered by a regulation of the European Pa rliament and of the

− Council on hor izontal cybersecur ity requirements f or products with digital elements. Theref ore, by wa y of

− deroga tion from this r ule, high-r isk AI syste ms that f all within the scope of this Regulation and are also qualified as

− im por tant and cr itical products with digital elements pursuant to a regulation of the European Parliament and of the

− Council on hor izontal cybersecur ity requirements f or products with digital elements and t o which the conf or mity

− assessment procedure based on inter nal control set out in an annex t o this Regulation applies, are subject to the

− conf or mity assessment provisions of a regulation of the European Parliament and of the Council on hor izontal

− cybersecur ity requirements f or products with digital elements insofar as the essential cybersecur ity requirements of

− that regulation are concer ned. In this case, f or all the other aspects covered by this Regulation the respective

− pro visions on conf or mity assessment based on inter nal control set out in an annex to this Regulation should apply .

− Building on the kno wledge and exper tise of ENISA on the cybersecur ity policy and tasks assigned t o ENISA under

− the Regulation (EU) 2019/881 of the European Parliament and of the Council (

− 37

− ), the Commission should cooperate

− with ENISA on issues related t o cybersecur ity of AI syste ms.

− 22/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− (

− 37

− ) Regulation (EU) 2019/881 of the European Parliament and of the Council of 17 Apr il 2019 on ENISA (the European Union Ag ency

− f or Cybersecur ity) and on inf or mation and communications tec hnology cybersecur ity cer tification and repealing Regulation

− (EU) No 526/2013 (Cybersecur ity A ct) (OJ L 151, 7.6.2019, p. 15).

− (79) It is appropr iate that a specif ic natural or lega l person, def ined as the pro vider , takes responsibility f or the placing on

− the marke t or the putting into ser vice of a high-r isk AI syste m, regardless of whether that natural or lega l person is

− the person who designed or developed the syste m.

− (80) As signat or ies t o the Uni te d Nations Convention on the Rights of Pe rsons with Disabilities, the Union and the

− Member Stat es are legally oblige d to prot ect persons with disabilities from discr imination and promot e their equality ,

− t o ensure that persons with disabilities hav e access, on an equal basis with others, to inf or mation and

− communications te chnologies and systems, and to ensure respect f or pr ivacy f or persons with disabilities. Given the

− gro wing impor tance and use of AI systems, the application of universal design pr inciples to all new te chnologi es and

− ser vices should ensure full and equal access f or ever y one potentially affect ed by or using AI t echnologies, including

− persons with disabilities, in a wa y that take s full account of their inherent dignity and diversity . It is theref ore

− essential that provid ers ensure full compliance with accessibility requirements, including Directive (EU) 2016/2102

− of the European Pa rliament and of the Council (

− 38

− ) and Directive (EU) 2019/882. Providers should ensure

− com pliance with these requirements by design. Theref ore, the necessar y measures should be integrat ed as much as

− possible into the design of the high-r isk AI syste m.

− (81) The provider should establish a sound quality manag ement syste m, ensure the accom plishment of the required

− conf or mity assessment procedure, draw up the relevant documentation and establish a robust post-mark et

− monito r ing syste m. Provi ders of high-r isk AI systems that are subject to oblig ations regarding quality management

− syste ms under relevant secto ral Uni on la w should have the possibility t o include the elements of the quality

− manag ement syste m provid ed f or in this Regulation as par t of the existing quality management system provid ed f or

− in that other sectoral Uni on law . The complement ar ity between this Regulation and existing secto ral Union la w

− should also be take n into account in future standardisation activities or guidance adop ted by the Commission. Public

− author ities which put into ser vice high-r isk AI syste ms f or their own use ma y adop t and im plement the r ules f or the

− quality managem ent syste m as par t of the quality managem ent syste m adop te d at a national or regional level, as

− appropr iat e, taking into account the specificities of the secto r and the comp etences and org anisation of the public

− author ity concer ned.

− (82) T o enable enf orcement of this Regulation and create a level pla ying field f or operat ors, and, taking into account the

− diff erent f or ms of making ava ilable of digital products, it is impor tant to ensure that, under all circumstances,

− a person established in the Uni on can provide author ities with all the necessar y inf or mation on the comp liance of an

− AI syste m. Theref ore, pr ior to making their AI systems ava ilable in the Uni on, provid ers established in third

− countr ies should, by wr itten mandat e, appoint an author ised representative established in the Union. This author ised

− representative pla ys a pivot al role in ensur ing the comp liance of the high-r isk AI systems placed on the marke t or

− put into ser vice in the Union by those providers who are not established in the Union and in ser ving as their contact

− person established in the Union.

− (83) In light of the nature and comp lexity of the value chain f or AI syste ms and in line with the New Legislative

− Framewo rk , it is essential to ensure lega l cer tainty and f acilitate the compliance with this Regulation. Theref ore, it is

− necessar y to clar ify the role and the specific obliga tions of relevant operat ors along that value chain, suc h as

− im por te rs and distr ibutors who ma y contr ibut e to the development of AI systems. In cer tain situations those

− operat ors could act in more than one role at the same time and should theref ore fulfil cumulatively all relevant

− oblig ations associate d with those roles. For exam ple, an operator could act as a distr ibutor and an imp or te r at the

− same time.

− (84) T o ensure legal cer tainty , it is necessar y to clar ify that, under cer tain specific conditions, any distr ibutor , imp or te r ,

− deplo y er or other third-par ty should be considered to be a provider of a high-r isk AI syste m and theref ore assume all

− the relevant obliga tions. This would be the case if that par ty puts its name or trademark on a high-r isk AI system

− already placed on the marke t or put into ser vice, without prejudice to contractual ar rang ements stipulating that the

− oblig ations are allocated other wise. This would also be the case if that par ty makes a substantial modifi cation to

− a high-r isk AI syste m that has already been placed on the market or has already been put into ser vice in a wa y that it

− remains a high-r isk AI syste m in accordance with this Regulation, or if it modif ies the inte nded pur pose of an AI

− syste m, including a general-purpo se AI system, which has not been classif ied as high-r isk and has already been

− placed on the mark et or put into ser vice, in a wa y that the AI syste m becomes a high-r isk AI syste m in accordance

− with this Regulation. Those provis ions should apply without prejudice to more specific pro visions established in

− cer tain Uni on har monisation legislation based on the New Legislative Framewo rk , tog ether with which this

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 23/144

− (

− 38

− ) Directive (EU) 2016/2102 of the European Parl iament and of the Council of 26 October 2016 on the accessibility of the websit es

− and mobile applications of public sector bodies (OJ L 327, 2.12.2016, p. 1).

− Regulation should apply . For example, Ar ticle 16(2) of Regulation (EU) 2017/745, establishing that cer tain ch ange s

− should not be considered to be modifi cations of a device that could affe ct its comp liance with the applicable

− requirements, should continue to apply to high-r isk AI syste ms that are medical devices within the meaning of that

− Regulation.

− (85) General-pur pose AI syste ms ma y be used as high-r isk AI syste ms by themselves or be comp onents of other high-r isk

− AI syste ms. Theref ore, due to their par ticular nature and in order to ensure a f air shar ing of responsibilities along the

− AI value ch ain, the provider s of such syste ms should, ir respective of whether they ma y be used as high-r isk AI

− syste ms as such by other providers or as compo nents of high-r isk AI syste ms and unless provid ed other wise under

− this Regulation, closely cooperate with the providers of the relevant high-r isk AI systems to enable their compliance

− with the relevant obligations under this Regulation and with the compet ent author ities established under this

− Regulation.

− (86) Where, under the conditions laid do wn in this Regulation, the provid er that initially placed the AI syste m on the

− mark et or put it into ser vice should no longe r be considered to be the provid er f or the pur poses of this Regulation,

− and when that provid er has not expressly excluded the chang e of the AI syste m into a high-r isk AI syste m, the

− f or mer provid er should nonetheless closely cooperate and make ava ilable the necessar y inf or mation and provid e the

− reasonably expecte d te chnical access and other assistance that are required f or the fulf ilment of the obligations set

− out in this Regulation, in par ticular regarding the comp liance with the conf or mity assessment of high-r isk AI

− syste ms.

− (87) In addition, where a high-r isk AI syste m that is a saf ety comp onent of a product which f alls within the scope of

− Uni on har monisation legislation based on the New Legislative Framework is not placed on the market or put into

− ser vice independently from the product, the product manufa cturer def ined in that legislation should comp ly with

− the obliga tions of the provider established in this Regulation and should, in par ticular , ensure that the AI system

− embedded in the final product comp lies with the requirements of this Regulation.

− (88) Along the AI value chain multiple par ties of te n supply AI syste ms, tools and ser vices but also compo nents or

− processes that are incor porate d b y the provid er into the AI syste m with var ious objectives, including the model

− training, model retraining, model te sting and evaluation, inte gration into sof tware, or other aspects of model

− development. Those par ties have an imp or tant role to pla y in the value ch ain to wards the provid er of the high-r isk

− AI syste m into which their AI systems, tools, ser vices, comp onents or processes are integrat ed, and should provide

− b y wr itt en agreement this provid er with the necessar y inf or mation, capabilities, tec hnical access and other assistance

− based on the generally ackno wledged stat e of the ar t, in order to enable the provid er to fully comply with the

− oblig ations set out in this Regulation, without com promising their ow n intellect ual proper ty r ights or trade secrets.

− (89) Third par ties making accessible to the public too ls, ser vices, processes, or AI comp onents other than

− g eneral-pur pose AI models, should not be mandat ed to comp ly with requirements targ eting the responsibilities

− along the AI value ch ain, in par ticular to wards the provid er that has used or integrat ed them, when those t ools,

− ser vices, processes, or AI comp onents are made accessible under a free and open-source licence. Developers of free

− and open-source tools, ser vices, processes, or AI comp onents other than g eneral-pur pose AI models should be

− encourag ed t o im plement widely adop ted documentation practices, suc h as model cards and data sheets, as a wa y to

− accelerat e inf or mation shar ing along the AI value chain, allowi ng the promotion of tr ustwor thy AI systems in the

− Uni on.

− (90) The Commission could develop and recommend voluntary model contractual term s between providers of high-r isk

− AI syste ms and third par ties that supply to ols, ser vices, comp onents or processes that are used or integrat ed in

− high-r isk AI systems, to f acilitate the cooperation along the value ch ain. When developing v oluntar y model

− contractual te r ms, the Commission should also tak e into account possible contractual requirements applicable in

− specific sectors or business cases.

− (91) Given the nature of AI systems and the r isk s to safety and fundamental r ights possibly associate d with their use,

− including as rega rds the need to ensure proper monitori ng of the perfor mance of an AI syste m in a real-life setting, it

− is appropr iate to set specific responsibilities f or deplo y ers. Deplo y ers should in par ticular tak e appropr iat e t echnical

− and organisational measures to ensure the y use high-r isk AI syste ms in accordance with the instr uctions of use and

− cer tain other obliga tions should be pro vided f or with regard to monitoring of the functioning of the AI systems and

− with regard to record-keeping, as appropr iate. Fur ther more, deplo y ers should ensure that the persons assigned to

− im plement the instr uctions f or use and human ove rsight as set out in this Regulation hav e the necessar y

− 24/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− com petence, in par ticular an adequate level of AI literacy , training and author ity to properly fulf il those tasks. Those

− oblig ations should be without prejudice to other deplo y er obliga tions in relation to high-r isk AI syste ms under

− Uni on or national la w .

− (92) This Regulation is without prejudice to obligations f or emplo y ers to inf or m or t o inf or m and consult w orkers or

− their representatives under Uni on or national la w and practice, including Directive 2002/14/EC of the European

− P arliament and of the Council (

− 39

− ), on decisions to put into ser vice or use AI syste ms. It remains necessar y to ensure

− inf or mation of w orkers and their representatives on the planned deplo yment of high-r isk AI syste ms at the

− w orkplace where the conditions f or those inf or mation or inf or mation and consultation obliga tions in other leg al

− instr uments are not fulfilled. Moreo ver , such inf or mation r ight is ancillar y and necessar y to the objective of

− prot ecting fundamental r ights that underlies this Regulation. Theref ore, an inf or mation requirement t o that effect

− should be laid do wn in this Regulation, without affe cting any existing r ights of work ers.

− (93) Whilst r isk s relate d to AI systems can result from the wa y suc h systems are designed, r isks can as well st em from

− ho w such AI systems are used. Deplo y ers of high-r isk AI syste m theref ore pla y a cr itical role in ensur ing that

− fundamental r ights are protect ed, compl ementing the obligati ons of the provider when developing the AI syste m.

− Deplo y ers are best placed t o understand ho w the high-r isk AI syste m will be used concretely and can theref ore

− identify potent ial signif icant r isks that were not f oreseen in the development phase, due to a more precise kno wledge

− of the context of use, the persons or groups of persons like ly to be affect ed, including vulnerable groups. Deplo yers

− of high-r isk AI syste ms list ed in an annex to this Regulation also pla y a cr itical role in inf or ming natural persons and

− should, when they mak e decisions or assist in making decisions related to natural persons, where applicable, inf or m

− the natural persons that the y are subject to the use of the high-r isk AI system. This inf or mation should include the

− inte nded pur pose and the type of decisions it makes. The deplo yer should also inf or m the natural persons about

− their r ight to an explanation provided under this Regulation. With regard to high-r isk AI syste ms used f or la w

− enf orcement pur poses, that obligation should be imp lemented in accordance with Ar ticle 13 of Directive (EU)

− 2016/680.

− (94) Any processing of biometr ic data in volved in the use of AI systems f or biometr ic identifica tion f or the pur pose of

− la w enf orcement needs to comp ly with Ar ticle 10 of Directive (EU) 2016/680, that allo ws suc h processing only

− where str ictly necessar y , subject to appropr iate safeguards f or the r ights and freedoms of the data subject, and where

− author ised by Union or Member Stat e law . Such use, when author ised, also needs to respect the pr inciples laid do wn

− in Ar ticle 4 (1) of Directive (EU) 2016/680 including law fulness, f air ness and transparency , pur pose limitation,

− accuracy and storag e limitation.

− (95) Without prejudice to applicable Uni on law , in par ticular Regulation (EU) 2016/679 and Directive (EU) 2016/680,

− consider ing the intr usive nature of post-remote biometr ic identifica tion systems, the use of post-remote biometr ic

− identification syste ms should be subject to saf eguards. P ost-remot e biometr ic identification systems should alwa ys be

− used in a wa y that is propor tionate , legitimate and str ictly necessar y , and thus targe te d, in te r ms of the individuals to

− be identified, the location, t emporal scope and based on a closed data set of leg ally acquired video f ootage. In any

− case, post-remote biometr ic identifica tion syste ms should not be used in the framework of law enf orcement to lead

− t o indiscr iminate sur veillance. The conditions f or post-remote biometr ic identifica tion should in any case not

− pro vide a basis to circumvent the conditions of the prohibition and str ict excep tions f or real time remote biometr ic

− identification.

− (96) In order to eff iciently ensure that fundamental r ights are protect ed, deplo y ers of high-r isk AI syste ms that are bodies

− gove r ned by public law , or pr ivate entities providing public ser vices and deplo y ers of cer tain high-r isk AI syste ms

− list ed in an annex to this Regulation, suc h as banking or insurance entities, should car r y out a fundamental r ights

− im pact assessment pr ior to putting it into use. Ser vices imp or tant f or individuals that are of public nature ma y also

− be provided by pr ivate entities. Pr ivat e entities pro viding suc h public ser vices are linked t o tasks in the public interest

− suc h as in the areas of education, healthcare, social ser vices, housing, administration of justice. The aim of the

− fundamental r ights impact assessment is f or the deplo y er to identify the specif ic r isks to the r ights of individuals or

− groups of individuals like ly to be affe cted , identify measures to be take n in the case of a mat er ialisation of those r isks.

− The impact assessment should be perf or med pr ior to deplo ying the high-r isk AI syste m, and should be updated

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 25/144

− (

− 39

− ) Directive 2002/14/EC of the European Parliament and of the Council of 11 March 2002 establishing a ge neral framew ork f or

− inf or ming and consulting em plo yee s in the European Community (OJ L 80, 23.3.2002, p. 29).

− when the deplo y er considers that any of the relevant f actors hav e ch anged. The im pact assessment should identify

− the deplo yer ’s relevant processes in which the high-r isk AI system will be used in line with its intended pur pose, and

− should include a descr iption of the per iod of time and frequency in which the syste m is intended to be used as well

− as of specific categor ies of natural persons and groups who are like ly to be affect ed in the specif ic context of use. The

− assessment should also include the identifi cation of specific r isks of har m likel y to hav e an im pact on the

− fundamental r ights of those persons or groups. While per f or ming this assessment, the deplo y er should tak e into

− account inf or mation relevant to a proper assessment of the imp act, including but not limit ed to the inf or mation

− given by the provid er of the high-r isk AI system in the instr uctions f or use. In light of the r isks identified, deplo yers

− should det er mine measures to be taken in the case of a materialis ation of those r isks , including f or exam ple

− gove r nance ar rangements in that specific context of use, suc h as ar rang ements f or human ove rsight according to the

− instr uctions of use or , complai nt handling and redress procedures, as they could be instr umental in mitigating r isks

− t o fundamental r ights in concrete use-cases. Af te r perfor ming that impact assessment, the deplo y er should notify the

− relevant marke t sur veillance author ity . Where appropr iate, to collect relevant inf or mation necessar y to perfo r m the

− im pact assessment, deplo yers of high-r isk AI syste m, in par ticular when AI systems are used in the public sect or ,

− could inv olve relevant stakeh olders, including the representatives of groups of persons like ly to be affect ed by the AI

− syste m, independent exper ts, and civil society organisations in conducting suc h im pact assessments and designing

− measures to be take n in the case of materi alisation of the r isks . The European Ar tif icial Intellig ence Off ice (AI Offi ce)

− should develop a t emplat e f or a questionnaire in order to f acilitate comp liance and reduce the administrative burden

− f or deplo y ers.

− (97) The notion of g eneral-pur pose AI models should be clearly defined and set apar t from the notion of AI systems to

− enable legal cer tainty . The def inition should be based on the ke y functional ch aracter istics of a general-pur pose AI

− model, in par ticular the generality and the capability t o com petently perf or m a wide rang e of distinct tasks. These

− models are typically trained on larg e amounts of data, through var ious methods, suc h as self-super vised,

− unsuper vised or reinf orcement lear ning. General-pur pose AI models ma y be placed on the market in var ious wa ys,

− including through librar ies, application programming interfaces (APIs), as direct download, or as physica l cop y .

− These models ma y be fur ther modif ied or fine -tuned into new models. Although AI models are essential

− com ponents of AI systems, they do not constitute AI systems on their own. AI models require the addition of fur ther

− com ponents, such as f or exam ple a user inte rface, t o become AI syste ms. AI models are typically integrat ed into and

− f or m par t of AI systems. This Regulation provid es specific r ules f or g eneral-pur pose AI models and f or

− g eneral-pur pose AI models that pose syste mic r isks , which should apply also when these models are integrat ed or

− f or m par t of an AI syste m. It should be understood that the obliga tions f or the providers of gene ral-pur pose AI

− models should apply once the g eneral-pur pose AI models are placed on the marke t. When the provid er of

− a g eneral-pur pose AI model inte grates an ow n model into its ow n AI syste m that is made ava ilable on the market or

− put into ser vice, that model should be considered to be placed on the marke t and, theref ore, the obliga tions in this

− Regulation f or models should continue to apply in addition t o those f or AI syste ms. The obligations laid do wn f or

− models should in an y case not apply when an own model is used f or purely inter nal processes that are not essential

− f or providing a product or a ser vice to third par ties and the r ights of natural persons are not affe cted . Consider ing

− their potent ial signifi cantly nega tive effe cts, the g eneral-pur pose AI models with syste mic r isk should alwa ys be

− subject to the relevant obliga tions under this Regulation. The definition should not cover AI models used bef ore their

− placing on the mark et f or the sole pur pose of researc h, development and prot otyping activities. This is without

− prejudice to the obligation to com ply with this Regulation when, f ollo wing suc h activities, a model is placed on the

− mark et.

− (98) Whereas the g enerality of a model could, inter alia, also be det er mined by a number of paramet ers, models with at

− least a billion of paramet ers and trained with a larg e amount of data using self-super vision at scale should be

− considered to displa y significant generality and to compet ently perf or m a wide rang e of distinctive tasks .

− (99) Larg e g enerative AI models are a typical exam ple f or a g eneral-pur pose AI model, given that they allow f or f lexible

− g eneration of cont ent, suc h as in the f or m of te xt, audio, imag es or video, that can readily accommodate a wide

− rang e of distinctive tasks.

− (100) When a g eneral-pur pose AI model is inte grated into or f or ms par t of an AI syste m, this syste m should be considered

− t o be g eneral-pur pose AI syste m when, due to this integration, this system has the capability to ser ve a var iety of

− pur poses. A general-purpo se AI system can be used directly , or it ma y be integrat ed into other AI syste ms.

− 26/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− (101) Provi ders of general-purpo se AI models hav e a par ticular role and responsibility along the AI value chain, as the

− models the y provide ma y f or m the basis f or a rang e of downstream syste ms, of te n provided b y downstream

− pro viders that necessitate a good understanding of the models and their capabilities, both t o enable the integration of

− suc h models into their products, and to fulfil their obligations under this or other regulations. Theref ore,

− propor tionate transparency measures should be laid down, including the drawi ng up and keeping up to date of

− documentation, and the provision of inf or mation on the general-pur pose AI model f or its usage b y the downstream

− pro viders. T echnical documentation should be prepared and kep t up to dat e by the g eneral-pur pose AI model

− pro vider f or the pur pose of making it ava ilable, upon request, t o the AI Offi ce and the national compet ent

− author ities. The minimal set of elements to be included in such documentation should be set out in specific annexe s

− t o this Regulation. The Commission should be empo wered to amend those annexe s by means of delegat ed acts in

− light of evolving t echnological developments.

− (102) Sof tware and data, including models, released under a free and open-source licence that allo ws them to be openly

− shared and where users can freely access, use, modify and redistr ibute them or modif ied versions thereof, can

− contr ibut e to research and inno vation in the mark et and can pro vide significant growth oppor tunities f or the Uni on

− economy . General-pur pose AI models released under free and open-source licences should be considered to ensure

− high levels of transparency and openness if their parameter s, including the weights, the inf or mation on the model

− arc hitecture, and the inf or mation on model usage are made publicly ava ilable. The licence should be considered to be

− free and open-source also when it allows users to r un, cop y , distr ibut e, study , ch ange and im prove sof tware and data,

− including models under the condition that the or iginal provider of the model is credited, the identical or compar able

− t er ms of distr ibution are respected.

− (103) Free and open-source AI comp onents covers the sof tware and data, including models and g eneral-pur pose AI

− models, tools, ser vices or processes of an AI syste m. Free and open-source AI components can be provided through

− diff erent ch annels, including their development on open repositor ies. For the pur poses of this Regulation, AI

− com ponents that are provid ed aga inst a pr ice or other wise monetised, including through the provision of t echnical

− suppor t or other ser vices, including through a sof tware platf or m, related to the AI comp onent, or the use of

− personal data f or reasons other than exclusively f or imp ro ving the secur ity , compatibility or interoperability of the

− sof tware, with the ex ception of transactions between microenter pr ises, should not benef it from the excep tions

− pro vided to free and open-source AI components. The fact of making AI components availa ble through open

− reposit or ies should not, in itself, constitute a monetisation.

− (104) The provid ers of g eneral-pur pose AI models that are released under a free and open-source licence, and whose

− paramet ers, including the weights, the inf or mation on the model archit ecture, and the inf or mation on model usage,

− are made publicly ava ilable should be subject to ex ceptions as rega rds the transparency-related requirements

− im posed on g eneral-pur pose AI models, unless they can be considered to present a systemic r isk, in which case the

− circumstance that the model is transparent and accompanied by an open-source license should not be considered to

− be a suffi cient reason to exclude compliance with the obligati ons under this Regulation. In any case, given that the

− release of general-pur pose AI models under free and open-source licence does not necessar ily reveal substantial

− inf or mation on the data set used f or the training or fine -tuning of the model and on how compliance of cop yr ight

− la w was thereby ensured, the excep tion provid ed f or g eneral-pur pose AI models from compliance with the

− transparency-relate d requirements should not concer n the obligati on to produce a summar y about the cont ent used

− f or model training and the obligation to put in place a policy to comply with Union cop yr ight law , in par ticular to

− identify and comply with the reser vation of r ights pursuant to Ar ticle 4(3) of Directive (EU) 2019/790 of the

− European Pa rliament and of the Council (

− 40

− ).

− (105) General-pur pose AI models, in par ticular larg e g enerative AI models, capable of g enerating text, imag es, and other

− cont ent, present unique innovation oppor tunities but also challeng es to ar tists, authors, and other creators and the

− wa y their creative content is create d, distr ibut ed, used and consumed. The development and training of suc h models

− require access to vast amounts of te xt, images, videos and other data. T ext and data mining te chniques ma y be used

− exte nsively in this context f or the retr ieval and analysis of suc h cont ent, which ma y be protect ed by cop yr ight and

− relate d r ights. Any use of cop yr ight protect ed cont ent requires the author isation of the r ightsholder concer ned

− unless relevant cop yr ight excep tions and limitations apply . Directive (EU) 2019/790 introduced ex ceptions and

− limitations allowing reproductions and extractions of wo rk s or other subject matter , f or the pur pose of te xt and data

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 27/144

− (

− 40

− ) Directive (EU) 2019/790 of the European Parliament and of the Council of 17 Apr il 2019 on cop yr ight and relat ed r ights in the

− Digital Sing le Market and amending Directives 96/9/EC and 2001/29/EC (OJ L 130, 17.5.2019, p. 92).

− mining, under cer tain conditions. Und er these r ules, r ightsholders ma y choose to reser ve their r ights over their

− w ork s or other subject matter to prevent te xt and data mining, unless this is done f or the pur poses of scientific

− researc h. Where the r ights to op t out has been expressly reser ved in an appropr iate manner , provid ers of

− g eneral-pur pose AI models need to obtain an author isation from r ightsholders if the y want to car r y out t ext and

− data mining ove r suc h w ork s.

− (106) Provi ders that place g eneral-pur pose AI models on the Union market should ensure com pliance with the relevant

− oblig ations in this Regulation. T o that end, providers of g eneral-pur pose AI models should put in place a policy to

− com ply with Uni on law on cop yr ight and related r ights, in par ticular to identify and com ply with the reser vation of

− r ights expressed b y r ightsholders pursuant to Ar ticle 4(3) of Directive (EU) 2019/790. Any provider placing

− a g eneral-pur pose AI model on the Uni on market should comply with this obligation, regard less of the jur isdiction

− in which the cop yr ight-relevant acts under pinning the training of those general-pur pose AI models take place. This

− is necessar y to ensure a level pla ying fie ld among provider s of gene ral-pur pose AI models where no provider should

− be able t o g ain a comp etitive advantage in the Union market by applying lower cop yr ight standards than those

− pro vided in the Uni on.

− (107) In order to increase transparency on the data that is used in the pre-training and training of general-pur pose AI

− models, including text and data protect ed by cop yr ight law , it is adequate that provid ers of suc h models draw up and

− mak e publicly a vailable a sufficiently detailed summar y of the content used f or training the g eneral-pur pose AI

− model. While taking into due account the need to protect trade secrets and conf idential business inf or mation, this

− summar y should be g enerally comp rehensive in its scope instead of te ch nically detailed to facilitat e par ties with

− legitimat e inte rests, including cop yr ight holders, t o ex ercise and enf orce their r ights under Uni on law , f or exam ple by

− listing the main data collections or sets that went into training the model, suc h as large pr ivate or public databases or

− data archives, and by providing a nar rative explanation about other data sources used. It is appropr iat e f or the AI

− Office to provid e a tem plate f or the summar y , which should be sim ple, effe ctive, and allo w the provid er t o provide

− the required summar y in nar rative f or m.

− (108) With rega rd to the obliga tions imp osed on providers of g eneral-pur pose AI models to put in place a policy to

− com ply with Uni on cop yr ight law and make publicly a vailable a summar y of the cont ent used f or the training, the AI

− Office should monitor whether the provider has fulf illed those obliga tions without ver ifying or proceeding to

− a wo rk -by-w ork assessment of the training data in ter ms of cop yr ight comp liance. This Regulation does not affect

− the enf orcement of cop yr ight r ules as provided f or under Uni on law .

− (109) Comp liance with the obliga tions applicable to the providers of g eneral-pur pose AI models should be commensurate

− and propor tionate to the type of model pro vider , ex cluding the need f or compliance f or persons who develop or use

− models f or non-profess ional or scientific research pur poses, who should never theless be encouraged to v oluntar ily

− com ply with these requirements. Without prejudice to Uni on cop yr ight law , comp liance with those obligations

− should take due account of the size of the provider and allow sim plif ied wa ys of compliance f or SMEs, including

− star t-ups, that should not represent an ex cessive cost and not discourage the use of suc h models. In the case of

− a modifi cation or fine -tuning of a model, the obligations f or provid ers of g eneral-pur pose AI models should be

− limit ed to that modifi cation or fine-tuning, f or exam ple by complementing the already existing t echnical

− documentation with inf or mation on the modifi cations, including new training data sources, as a means to com ply

− with the value ch ain oblig ations provided in this Regulation.

− (110) General-pur pose AI models could pose systemic r isks which include, but are not limit ed t o, any actual or reasonably

− f oreseeable nega tive effe cts in relation to major accidents, disr up tions of cr itical sectors and ser ious consequences to

− public health and saf ety ; any actual or reasonably f oreseeable negative effe cts on democratic processes, public and

− economic secur ity ; the dissemination of illega l, false, or discr iminat or y cont ent. Systemic r isks should be understood

− t o increase with model capabilities and model reach, can ar ise along the entire lifecy cle of the model, and are

− inf luenced by conditions of misuse, model reliability , model f air ness and model secur ity , the level of autonom y of

− 28/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− the model, its access to too ls, novel or combined modalities, release and distr ibution strategies, the potential to

− remo ve guardrails and other f actor s. In par ticular , inte r national approac hes have so f ar identifie d the need to pa y

− atte ntion to r isks from pote ntial intent ional misuse or uninte nded issues of control relating to alignment with

− human intent; chemical, biological, radiological, and nuclear r isks , suc h as the wa ys in which bar r iers to entr y can be

− lo wered, including f or weapons development, design acquisition, or use; offe nsive cyber capabilities, suc h as the

− wa ys in vulnerability discove r y , exploitation, or operational use can be enabled; the eff ects of interac tion and to ol

− use, including f or exam ple the capacity to control physical systems and interfere with cr itical infrastr ucture; r isks

− from models of making copies of themselves or ‘self-replicating’ or training other models; the wa ys in which models

− can give r ise to har mful bias and discr imination with r isks to individuals, communities or societies; the facili tation of

− disinf or mation or har ming pr ivacy with threats to democratic values and human r ights; r isk that a par ticular event

− could lead to a ch ain reaction with considerable nega tive effe cts that could affe ct up to an entire city , an entire

− domain activity or an entire community .

− (111) It is appropr iate to establish a methodology f or the classif ication of g eneral-pur pose AI models as g eneral-pur pose

− AI model with systemic r isks . Since syste mic r isks result from par ticularly high capabilities, a g eneral-pur pose AI

− model should be considered to present systemic r isks if it has high-impact capabilities, evaluate d on the basis of

− appropr iat e te chnical to ols and methodologies, or significant impact on the internal market due to its reac h.

− High-im pact capabilities in general-pur pose AI models means capabilities that matc h or ex ceed the capabilities

− recorded in the most advanced general-pur pose AI models. The full range of capabilities in a model could be bett er

− understo od af te r its placing on the marke t or when deplo y ers interact with the model. A ccording to the state of the

− ar t at the time of entr y into f orce of this Regulation, the cumulative amount of comp utation used f or the training of

− the general-pur pose AI model measured in f loating point operations is one of the relevant appro ximations f or model

− capabilities. The cumulative amount of comp utation used f or training includes the computation used across the

− activities and methods that are intende d to enhance the capabilities of the model pr ior to deplo yment, suc h as

− pre-training, synthetic data generation and fine -tuning. Theref ore, an initial threshold of f loating point operations

− should be set, which , if met by a g eneral-pur pose AI model, leads to a presump tion that the model is

− a general-pur pose AI model with systemic r isks . This threshold should be adjuste d ove r time to ref lect te chnological

− and industr ial chang es, suc h as algor ithmic im provements or increased hardware effi ciency , and should be

− supplement ed with bench marks and indicator s f or model capability . T o inf or m this, the AI Offi ce should enga ge

− with the scientific community , industr y , civil society and other exper ts. Thresholds, as well as too ls and bench mark s

− f or the assessment of high-imp act capabilities, should be strong predict ors of g enerality , its capabilities and

− associat ed systemic r isk of general-pur pose AI models, and could take into account the wa y the model will be placed

− on the market or the number of users it ma y affect. T o com plement this system, there should be a possibility f or the

− Commission to take individual decisions designating a g eneral-pur pose AI model as a g eneral-pur pose AI model

− with systemic r isk if it is f ound that suc h model has capabilities or an im pact equivalent to those capture d by the set

− threshold. That decision should be take n on the basis of an overa ll assessment of the cr ite r ia f or the designation of

− a g eneral-pur pose AI model with systemic r isk set out in an annex to this Regulation, such as quality or size of the

− training data set, number of business and end users, its input and output modalities, its level of autonom y and

− scalability , or the to ols it has access to. Upon a reasoned request of a provider whose model has been designated as

− a g eneral-pur pose AI model with systemic r isk, the Commission should take the request into account and ma y

− decide to reassess whether the general-pur pose AI model can still be considered to present syste mic r isks.

− (112) It is also necessar y to clar ify a procedure f or the classific ation of a g eneral-pur pose AI model with systemic r isks .

− A g eneral-pur pose AI model that meets the applicable threshold f or high-imp act capabilities should be presumed to

− be a g eneral-pur pose AI models with syste mic r isk. The provid er should notify the AI Offi ce at the latest tw o weeks

− af te r the requirements are met or it becomes kno wn that a general-pur pose AI model will meet the requirements

− that lead to the presump tion. This is especially relevant in relation to the threshold of f loating point operations

− because training of gene ral-pur pose AI models take s considerable planning which includes the upfront allocation of

− com pute resources and, theref ore, provid ers of g eneral-pur pose AI models are able to kno w if their model w ould

− meet the threshold bef ore the training is complet ed. In the cont ext of that notif ication, the provid er should be able to

− demonstrat e that, because of its specif ic ch aracter istics, a g eneral-pur pose AI model ex ceptionally does not present

− syste mic r isks , and that it thus should not be classified as a g eneral-pur pose AI model with syste mic r isk s. That

− inf or mation is valuable f or the AI Off ice to anticipate the placing on the market of general-pur pose AI models with

− syste mic r isks and the provider s can star t to engag e with the AI Office early on. That inf or mation is especially

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 29/144

− im por tant with regard to general-pur pose AI models that are planned t o be released as open-source, given that, af ter

− the open-source model release, necessar y measures to ensure compliance with the obliga tions under this Regulation

− ma y be more diffic ult to imp lement.

− (113) If the Commission becomes awar e of the f act that a general-pur pose AI model meets the requirements to classify as

− a g eneral-pur pose AI model with syste mic r isk, which previously had either not been kno wn or of which the

− relevant provider has f ailed to notify the Commission, the Commission should be empo wered to designate it so.

− A system of qualified aler ts should ensure that the AI Offi ce is made aw are by the scientifi c panel of g eneral-pur pose

− AI models that should possibly be classified as g eneral-pur pose AI models with syste mic r isk, in addition to the

− monito r ing activities of the AI Off ice.

− (114) The pro viders of g eneral-pur pose AI models presenting syste mic r isks should be subject, in addition t o the

− oblig ations provid ed f or provid ers of g eneral-pur pose AI models, to oblig ations aimed at identifying and mitig ating

− those r isks and ensur ing an adequate level of cybersecur ity protection, regardless of whether it is provided as

− a standalone model or embedded in an AI system or a product. T o achi eve those objectives, this Regulation should

− require provid ers to perfo r m the necessar y model evaluations, in par ticular pr ior t o its fir st placing on the mark et,

− including conducting and documenting adversar ial t esting of models, also, as appropr iate, through inter nal or

− independent exte r nal testing. In addition, provider s of g eneral-pur pose AI models with systemic r isks should

− continuously assess and mitig ate systemic r isks, including f or exam ple by putting in place r isk -management policies,

− suc h as accountability and gover nance processes, imp lementing post-mark et monitoring, taking appropr iate

− measures along the entire model’s lif ecy cle and cooperating with relevant actor s along the AI value chai n.

− (115) Provi ders of general-pur pose AI models with systemic r isks should assess and mitiga te possible systemic r isks . If,

− despite eff or ts t o identify and prevent r isks related to a g eneral-pur pose AI model that ma y present syste mic r isks ,

− the development or use of the model causes a ser ious incident, the general-pur pose AI model provid er should

− without undue dela y keep trac k of the incident and repor t any relevant inf or mation and possible cor rective measures

− t o the Commission and national compet ent author ities. Fur ther more, provider s should ensure an adequate level of

− cybersecur ity prot ection f or the model and its physical infrastr ucture, if appropr iate, along the entire model lifecy cle.

− Cybersecur ity protection related to systemic r isks associated with malicious use or attacks should duly consider

− accidental model leakage, unauthor ised releases, circum vention of saf ety measures, and defe nce aga inst cyberattacks,

− unauthor ised access or model thef t. That prot ection could be facili tated by secur ing model weights, algor ithms,

− ser vers, and data sets, suc h as through operational secur ity measures f or inf or mation secur ity , specif ic cybersecur ity

− policies, adequat e t echnical and established solutions, and cyber and phys ical access controls, appropr iate to the

− relevant circumstances and the r isks inv olved.

− (116) The AI Office should encourage and f acilitate the drawing up, review and adap tation of codes of practice, taking into

− account inter national approaches. All pro viders of gene ral-pur pose AI models could be invit ed to par ticipate. T o

− ensure that the codes of practice ref lect the stat e of the ar t and duly take into account a diverse set of perspectives,

− the AI Off ice should collaborate with relevant national comp etent author ities, and could, where appropr iate, consult

− with civil society org anisations and other relevant stakeholders and exper ts, including the Scientif ic Panel, f or the

− dra wing up of suc h codes. Codes of practice should co ver oblig ations f or provider s of g eneral-pur pose AI models

− and of g eneral-pur pose AI models presenting systemic r isks. In addition, as regard s systemic r isks , codes of practice

− should help to establish a r isk tax onomy of the type and nature of the syste mic r isks at Union level, including their

− sources. Codes of practice should also be f ocused on specific r isk assessment and mitiga tion measures.

− (117) The codes of practice should represent a central too l f or the proper compliance with the obliga tions provided f or

− under this Regulation f or provider s of general-pur pose AI models. Provi ders should be able to rely on codes of

− practice to demonstrate compliance with the obliga tions. By means of imp lementing acts, the Commission ma y

− decide to approve a code of practice and give it a g eneral validity within the Uni on, or , alt er natively , to provide

− common r ules f or the imp lementation of the relevant obliga tions, if, by the time this Regulation becomes applicable,

− a code of practice cannot be final ised or is not deemed adequate by the AI Office. Once a har monised standard is

− 30/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− published and assessed as suitable to cover the relevant oblig ations by the AI Office, comp liance with a European

− har monised standard should grant providers the presump tion of conf or mity . Provi ders of general-pur pose AI

− models should fur ther more be able to demonstrate comp liance using alter native adequate means, if codes of practice

− or har monised standards are not a vailable, or they choose not to rely on those.

− (118) This Regulation regulate s AI systems and AI models b y imposing cer tain requirements and obliga tions f or relevant

− mark et actor s that are placing them on the market, putting into ser vice or use in the Union, thereby complementing

− oblig ations f or provid ers of inter mediar y ser vices that embed suc h systems or models into their ser vices regulate d by

− Regulation (EU) 2022/2065. T o the extent that suc h syste ms or models are embedded into designated ver y larg e

− online platf or ms or ver y larg e online search engines, the y are subject t o the r isk -manag ement framew ork provid ed

− f or in Regulation (EU) 2022/2065. Consequently , the cor responding obliga tions of this Regulation should be

− presumed to be fulfilled, unless significant syste mic r isks not co vered by Regulation (EU) 2022/2065 emerg e and are

− identified in suc h models. Within this framework, pro viders of ver y larg e online platf or ms and ver y larg e online

− searc h engines are oblige d to assess potential systemic r isks st emming from the design, functioning and use of their

− ser vices, including how the design of algor ithmic systems used in the ser vice ma y contr ibut e to such r isks, as well as

− syste mic r isks st emming from potent ial misuses. Those provid ers are also oblige d to take appropr iate mitig ating

− measures in obser vance of fundamental r ights.

− (119) Consider ing the quick pace of inno vation and the te chnologi cal evolution of digital ser vices in scope of different

− instr uments of Uni on law in par ticular hav ing in mind the usage and the percep tion of their recipients, the AI

− syste ms subject to this Regulation ma y be provid ed as interm ediar y ser vices or par ts thereof within the meaning of

− Regulation (EU) 2022/2065, which should be inter prete d in a te chnology-neutral manner . For exam ple, AI syste ms

− ma y be used to provide online search engines, in par ticular , to the exte nt that an AI syste m suc h as an online ch atbot

− perf or ms search es of, in pr inciple, all websites, then incor porates the results into its existing knowle dge and uses the

− updat ed knowledg e to g enerate a sing le output that combines different sources of inf or mation.

− (120) Fur ther more, obliga tions placed on provid ers and deplo yers of cer tain AI syste ms in this Regulation to enable the

− det ection and disclosure that the outputs of those syste ms are ar tif icially generat ed or manipulated are par ticularly

− relevant to f acilitate the effe ctive imp lementation of Regulation (EU) 2022/2065. This applies in par ticular as regard s

− the obliga tions of provid ers of ver y larg e online platf or ms or ver y large online search engines to identify and

− mitig ate syste mic r isks that ma y ar ise from the dissemination of content that has been ar tif icially g enerated or

− manipulat ed, in par ticular r isk of the actual or f oreseeable negati ve effe cts on democratic processes, civic discourse

− and elect oral processes, including through disinf or mation.

− (121) Standardisation should pla y a k ey role to provid e tec hnical solutions to provider s to ensure comp liance with this

− Regulation, in line with the state of the ar t, to promote inno vation as well as comp etitiveness and gro wth in the

− sing le market. Comp liance with har monised standards as defined in Ar ticle 2, point (1)(c), of Regulation (EU)

− No 1025/2012 of the European Parliament and of the Council (

− 41

− ), which are nor mally expected to ref lect the state

− of the ar t, should be a means f or providers to demonstrate conf or mity with the requirements of this Regulation.

− A balanced representation of inte rests in volving all relevant stak eholders in the development of standards, in

− par ticular SMEs, consumer organisations and envir onmental and social stakeholders in accordance with Ar ticles 5

− and 6 of Regulation (EU) No 1025/2012 should theref ore be encouraged. In order to f acilitate compliance, the

− standardisation requests should be issued by the Commission without undue dela y . When prepar ing the

− standardisation request, the Commission should consult the advisor y f or um and the Board in order to collect

− relevant exper tise. However , in the absence of relevant refere nces to har monised standards, the Commission should

− be able to establish, via implementing acts, and af te r consultation of the advisor y f or um, common specif ications f or

− cer tain requirements under this Regulation. The common specif ication should be an ex ceptional f all back solution to

− f acilitate the provid er ’s obliga tion to comp ly with the requirements of this Regulation, when the standardisation

− request has not been accept ed b y any of the European standardisation org anisations, or when the relevant

− har monised standards insufficiently address fundamental r ights concer ns, or when the har monised standards do not

− com ply with the request, or when there are dela ys in the adoption of an appropr iate har monised standard. Where

− suc h a dela y in the adoption of a har monised standard is due t o the tec hnical complexity of that standard, this should

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 31/144

− (

− 41

− ) Regulation (EU) No 1025/2012 of the European Parliament and of the Council of 25 October 2012 on European standardisation,

− amending Council Directives 89/686/EEC and 93/15/EEC and Directives 94/9/EC, 94/25/EC, 95/16/EC, 97/23/EC, 98/34/EC,

− 2004/22/EC, 2007/23/EC, 2009/23/EC and 2009/105/EC of the European Parliament and of the Council and repealing Council

− Decision 87/95/EEC and Decision No 1673/2006/EC of the European Parliament and of the Council (OJ L 316, 14.11.2012, p. 12).

− be considered by the Commission bef ore contem plating the establishment of common specifications. When

− developing common specific ations, the Commission is encourage d to cooperate with international par tners and

− inte r national standardisation bodies.

− (122) It is appropr iate that, without prejudice to the use of har monised standards and common specifications, providers of

− a high-r isk AI syste m that has been trained and te sted on data ref lecting the specific geographical, behavi oural,

− cont extual or functional setting within which the AI system is inte nded to be used, should be presumed t o comp ly

− with the relevant measure provided f or under the requirement on data gove r nance set out in this Regulation.

− Without prejudice to the requirements related to robustness and accuracy set out in this Regulation, in accordance

− with Ar ticle 54(3) of Regulation (EU) 2019/881, high-r isk AI syste ms that hav e been cer tif ied or f or which

− a statement of conf or mity has been issued under a cybersecur ity scheme pursuant to that Regulation and the

− refe rences of which hav e been published in the Off icial Jour nal of the Eur opean Union should be presumed to comply

− with the cybersecur ity requirement of this Regulation in so f ar as the cybersecur ity cer tif icate or statem ent of

− conf or mity or par ts thereof cover the cybersecur ity requirement of this Regulation. This remains without prejudice

− t o the v oluntar y nature of that cybersecur ity scheme.

− (123) In order to ensure a high level of tr ustw or thiness of high-r isk AI systems, those systems should be subject to

− a conf or mity assessment pr ior to their placing on the marke t or putting into ser vice.

− (124) It is appropr iate that, in order to minimise the burden on operat ors and av oid an y possible duplication, f or high-r isk

− AI syste ms related to products which are cover ed by existing Union har monisation legislation based on the New

− Legislative Framewo rk , the compliance of those AI systems with the requirements of this Regulation should be

− assessed as par t of the conf or mity assessment already provided f or in that law . The applicability of the requirements

− of this Regulation should thus not affect the specif ic logic, methodology or g eneral str ucture of conf or mity

− assessment under the relevant Uni on har monisation legislation.

− (125) Given the comp lexity of high-r isk AI systems and the r isks that are associate d with them, it is impor tant to develop

− an adequate conf or mity assessment procedure f or high-r isk AI syste ms inv olving notified bodies, so-called third

− par ty conf or mity assessment. However , given the cur rent exper ience of professional pre-market cer tif iers in the fie ld

− of product saf ety and the different nature of r isk s inv olved, it is appropr iate to limit, at least in an initial phase of

− application of this Regulation, the scope of application of third-par ty conf or mity assessment f or high-r isk AI

− syste ms other than those related to products. Theref ore, the conf or mity assessment of suc h syste ms should be

− car r ied out as a g eneral r ule by the provider under its own responsibility , with the only excep tion of AI syste ms

− inte nded to be used f or biometr ics.

− (126) In order t o car r y out third-par ty conf or mity assessments when so required, notified bodies should be notif ied under

− this Regulation by the national comp etent author ities, provided that the y comp ly with a set of requirements, in

− par ticular on independence, compet ence, absence of conf licts of interests and suitable cybersecur ity requirements.

− Notification of those bodies should be sent by national compet ent author ities t o the Commission and the other

− Member Stat es by means of the electronic notif ication to ol developed and managed b y the Commission pursuant to

− Ar ticle R23 of Annex I t o Decision No 768/2008/EC.

− (127) In line with Uni on commitments under the W orld T rade Organization Agreement on T ec hnical Bar r iers to T rade, it is

− adequat e to f acilitate the mutual recognition of conf or mity assessment results produced by comp etent conf or mity

− assessment bodies, independent of the ter r itory in which the y are established, provid ed that those conf or mity

− assessment bodies established under the law of a third countr y meet the applicable requirements of this Regulation

− and the Union has concluded an agreement to that exte nt. In this context, the Commission should actively explore

− possible inte r national instr uments f or that pur pose and in par ticular pursue the conclusion of mutual recognition

− agreements with third countr ies.

− (128) In line with the commonly established notion of substantial modif ication f or products regulate d by Uni on

− har monisation legislation, it is appropr iate that whenever a chang e occurs which ma y affe ct the comp liance of

− a high-r isk AI system with this Regulation (e.g. ch ange of operating syste m or sof tware arch itec ture), or when the

− inte nded pur pose of the syste m ch ange s, that AI syste m should be considered to be a new AI system which should

− undergo a new conf or mity assessment. How ever , ch anges occur r ing to the algor ithm and the perf or mance of AI

− syste ms which continue to ‘lear n’ af ter being placed on the market or put into ser vice, namely automat ically

− adap ting how functions are car r ied out, should not constitute a substantial modifi cation, provid ed that those

− ch ange s hav e been pre-dete r mined by the provider and assessed at the moment of the conf or mity assessment.

− 32/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− (129) High-r isk AI systems should bear the CE marking to indicate their conf or mity with this Regulation so that they can

− move freely within the intern al mark et. For high-r isk AI systems embedded in a product, a physica l CE marking

− should be affixed, and ma y be comp lemented by a digital CE marking. For high-r isk AI systems only provid ed

− digitally , a digital CE marking should be used. Member States should not create unjustified obstacles t o the placing

− on the marke t or the putting into ser vice of high-r isk AI systems that comp ly with the requirements laid down in

− this Regulation and bear the CE marking.

− (130) Und er cer tain conditions, rapid a vailability of inno vative tec hnologies ma y be cr ucial f or health and saf ety of

− persons, the prot ection of the en vironment and climate ch ange and f or society as a whole. It is thus appropr iat e that

− under ex ceptional reasons of public secur ity or prot ection of life and health of natural persons, envir onmental

− prot ection and the prot ection of ke y industr ial and infrastr uctural assets, marke t sur veillance author ities could

− author ise the placing on the market or the putting into ser vice of AI systems which hav e not undergone

− a conf or mity assessment. In duly justified situations, as provided f or in this Regulation, law enf orcement author ities

− or civil prot ection author ities ma y put a specific high-r isk AI syste m into ser vice without the author isation of the

− mark et sur veillance author ity , provided that suc h author isation is request ed dur ing or af ter the use without undue

− dela y .

− (131) In order to f acilitate the wo rk of the Commission and the Member States in the AI field as well as to increase the

− transparency to wards the public, provider s of high-r isk AI syste ms other than those related to products f alling within

− the scope of relevant existing Union har monisation legislation, as well as providers who consider that an AI system

− list ed in the high-r isk use cases in an annex to this Regulation is not high-r isk on the basis of a deroga tion, should be

− required to register themselves and inf or mation about their AI syste m in an EU database, to be established and

− manag ed b y the Commission. Bef ore using an AI syste m listed in the high-r isk use cases in an annex to this

− Regulation, deplo y ers of high-r isk AI systems that are public author ities, agencies or bodies, should regist er

− themselves in suc h database and select the syste m that they envisag e t o use. Other deplo y ers should be entitled to do

− so v oluntar ily . This section of the EU database should be publicly accessible, free of charg e, the inf or mation should

− be easily navig able, understandable and machine-r eadable. The EU database should also be user -fr iendly , f or exam ple

− b y providing search functionalities, including through ke ywo rds, allowi ng the g eneral public t o find relevant

− inf or mation to be submitted upon the registration of high-r isk AI syste ms and on the use case of high-r isk AI

− syste ms, set out in an annex to this Regulation, to which the high-r isk AI systems cor respond. Any substantial

− modification of high-r isk AI systems should also be registered in the EU database. For high-r isk AI syste ms in the

− area of law enf orcement, migration, asylum and border control managem ent, the registration obliga tions should be

− fulfilled in a secure non-public section of the EU database. A ccess to the secure non-public section should be str ictly

− limit ed to the Commission as well as to market sur veillance author ities with regard to their national section of that

− database. High-r isk AI systems in the area of cr itical infrastr ucture should only be register ed at national level. The

− Commission should be the controller of the EU database, in accordance with Regulation (EU) 2018/1725. In order

− t o ensure the full functionality of the EU database, when deplo yed, the procedure f or setting the database should

− include the development of functional specif ications b y the Commission and an independent audit repor t. The

− Commission should take into account cybersecur ity r isks when car r ying out its tasks as data controller on the EU

− database. In order t o maximise the a vailability and use of the EU database by the public, the EU database, including

− the inf or mation made ava ilable through it, should comply with requirements under the Directive (EU) 2019/882.

− (132) Cer tain AI systems inte nded to interac t with natural persons or to g enerate cont ent ma y pose specific r isk s of

− im personation or decept ion ir respective of whether they qualify as high-r isk or not. In cer tain circumstances, the use

− of these systems should theref ore be subject to specif ic transparency obligations without prejudice to the

− requirements and obligations f or high-r isk AI systems and subject to target ed ex ceptions to take into account the

− special need of la w enf orcement. In par ticular , natural persons should be notified that they are inte racting with an AI

− syste m, unless this is obvious from the point of view of a natural person who is reasonably well-inf or med, obser vant

− and circumspect taking into account the circumstances and the context of use. When im plementing that obligation,

− the ch aracter istics of natural persons belonging to vulnerable groups due to their age or disability should be take n

− into account to the exte nt the AI system is intended to interact with those groups as well. Moreove r , natural persons

− should be notified when the y are exposed to AI systems that, by processing their biometr ic data, can identify or inf er

− the emotions or inte ntions of those persons or assign them to specific cate gor ies. Such specif ic categor ies can relate

− t o aspects suc h as sex, age , hair colour , e ye colour , tattoos, personal traits, ethnic or igin, personal pref erences and

− inte rests. Such inf or mation and notif ications should be provided in accessible f or mats f or persons with disabilities.

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 33/144

− (133) A var iety of AI systems can g enerate larg e quantities of synthetic cont ent that becomes increasing ly hard f or humans

− t o distinguish from human-generat ed and authentic cont ent. The wide availa bility and increasing capabilities of

− those systems hav e a significant impact on the integr ity and tr ust in the inf or mation ecosystem, raising new r isks of

− misinf or mation and manipulation at scale, fraud, im personation and consumer decep tion. In light of those im pacts,

− the fast t echnological pace and the need f or new methods and tec hniques t o trace or igin of inf or mation, it is

− appropr iat e to require providers of those systems to embed tec hnical solutions that enable marking in a machi ne

− readable f or mat and detect ion that the output has been g enerated or manipulate d b y an AI syste m and not a human.

− Such t echniques and methods should be sufficiently reliable, inte roperable, effe ctive and robust as far as this is

− t echnically feasible, taking into account ava ilable tec hniques or a combination of such te ch niques, suc h as

− wat er marks, metadata identifica tions, cr ypt ographic methods f or provin g prove nance and authenticity of cont ent,

− logging methods, finger pr ints or other te ch niques, as ma y be appropr iat e. When imp lementing this obligation,

− pro viders should also take into account the specif icities and the limitations of the different types of cont ent and the

− relevant t echnological and mark et developments in the fie ld, as ref lected in the generally ackno wledged state of the

− ar t. Such tec hniques and methods can be im plemented at the level of the AI syste m or at the level of the AI model,

− including g eneral-pur pose AI models generating content, thereby facilitating fulfilment of this obliga tion by the

− do wnstream provid er of the AI system. T o remain propor tionate , it is appropr iate to en visage that this marking

− oblig ation should not cover AI syste ms perf or ming pr imar ily an assistive function f or standard editing or AI syste ms

− not substantially alter ing the input data provided b y the deplo y er or the semantics thereof.

− (134) Fur ther to the te chnical solutions emp lo y ed b y the provider s of the AI system, deplo yers who use an AI syste m to

− g enerate or manipulate imag e, audio or video content that appreciably resembles existing persons, objects, places,

− entities or events and would f alsely appear to a person to be authentic or tr uthful (deep f ak es), should also clearly

− and distinguishably disclose that the cont ent has been ar tificially create d or manipulated by labelling the AI output

− according ly and disclosing its ar tif icial or igin. Compliance with this transparency obligation should not be

− inte r preted as indicating that the use of the AI syste m or its output impedes the r ight to freedom of expression and

− the r ight to freedom of the ar ts and sciences guarante ed in the Char ter , in par ticular where the content is par t of an

− evidently creative, satir ical, ar tistic, fi ctional or analogous work or programme, subject to appropr iate safeguards f or

− the r ights and freedoms of third par ties. In those cases, the transparency obliga tion f or deep f ak es set out in this

− Regulation is limited to disclosure of the exist ence of suc h g enerated or manipulate d cont ent in an appropr iat e

− manner that does not hamper the displa y or enjo yment of the w ork , including its nor mal exploitation and use, while

− maintaining the utility and quality of the w ork . In addition, it is also appropr iate t o envisag e a similar disclosure

− oblig ation in relation to AI-g enerated or manipulate d te xt to the extent it is published with the pur pose of inf or ming

− the public on matters of public interest unless the AI-g enerated content has undergone a process of human review or

− edit or ial control and a natural or leg al person holds editor ial responsibility f or the publication of the content.

− (135) Without prejudice to the mandat or y nature and full applicability of the transparency oblig ations, the Commission

− ma y also encourage and f acilitate the dra wing up of codes of practice at Union level to f acilitate the eff ective

− im plementation of the obliga tions regarding the det ection and labelling of ar tif icially g enerated or manipulated

− cont ent, including to suppor t practical ar rang ements f or making, as appropr iate, the det ection mec hanisms

− accessible and f acilitating cooperation with other actor s along the value chain, disseminating content or chec king its

− authenticity and provenance t o enable the public to effe ctively distinguish AI-g enerated cont ent.

− (136) The obliga tions placed on provid ers and deplo yers of cer tain AI syste ms in this Regulation t o enable the detection

− and disclosure that the outputs of those systems are ar tif icially generat ed or manipulat ed are par ticularly relevant to

− f acilitate the eff ective imp lementation of Regulation (EU) 2022/2065. This applies in par ticular as rega rds the

− oblig ations of providers of ver y larg e online platf or ms or ver y larg e online search engines to identify and mitigat e

− syste mic r isks that ma y ar ise from the dissemination of cont ent that has been ar tif icially g enerated or manipulate d,

− in par ticular the r isk of the actual or f oreseeable nega tive effe cts on democratic processes, civic discourse and

− elect oral processes, including through disinf or mation. The requirement to label content g enerated by AI syste ms

− under this Regulation is without prejudice to the obliga tion in Ar ticle 16(6) of Regulation (EU) 2022/2065 f or

− pro viders of hosting ser vices to process notices on illegal content received pursuant to Ar ticle 16(1) of that

− Regulation and should not inf luence the assessment and the decision on the illegality of the specif ic content. That

− assessment should be perform ed solely with reference to the r ules gover ning the lega lity of the cont ent.

− 34/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− (137) Comp liance with the transparency obliga tions f or the AI syste ms cover ed by this Regulation should not be

− inte r preted as indicating that the use of the AI syste m or its output is law ful under this Regulation or other Uni on

− and Member Stat e la w and should be without prejudice to other transparency obliga tions f or deplo yers of AI syste ms

− laid do wn in Union or national law .

− (138) AI is a rapidly developing f amily of tec hnologies that requires regulatory ove rsight and a safe and controlled space

− f or exper imentation, while ensur ing responsible innovation and integration of appropr iate safegua rds and r isk

− mitig ation measures. T o ensure a lega l framew ork that promotes innovation, is future-proof and resilient to

− disr up tion, Member States should ensure that their national compet ent author ities establish at least one AI

− regulat or y sandbo x at national level to f acilitate the development and te sting of inno vative AI syste ms under str ict

− regulat or y oversight bef ore these systems are placed on the marke t or other wise put into ser vice. Member States

− could also fulf il this obligation through par ticipating in already existing regulator y sandbo xe s or establishing jointly

− a sandbo x with one or more Member States’ compet ent author ities, insofar as this par ticipation provid es equivalent

− level of national coverag e f or the par ticipating Member Stat es. AI regulator y sandbo xe s could be established in

− phys ical, digital or hybri d f or m and ma y accommodate physica l as well as digital products. Establishing author ities

− should also ensure that the AI regulatory sandbo xe s have the adequate resources f or their functioning, including

− financial and human resources.

− (139) The objectives of the AI regulato r y sandbo xes should be t o f oster AI inno vation by establishing a controlled

− exper imentation and te sting environment in the development and pre-marketing phase with a view to ensur ing

− com pliance of the inno vative AI systems with this Regulation and other relevant Union and national law . Moreove r ,

− the AI regulato r y sandbo xes should aim t o enhance leg al cer tainty f or inno vators and the compet ent author ities’

− o versight and understanding of the oppor tunities, emerging r isks and the im pacts of AI use, to f acilitate regulator y

− lear ning f or author ities and under takings, including with a view t o future adap tions of the lega l framew ork, to

− suppor t cooperation and the shar ing of best practices with the author ities in volved in the AI regulator y sandbo x,

− and t o accelerate access to markets, including by removing bar r iers f or SMEs, including star t-ups. AI regulator y

− sandbo xe s should be widely ava ilable throughout the Uni on, and par ticular attention should be given to their

− accessibility f or SMEs, including star t-ups. The par ticipation in the AI regulato r y sandbo x should f ocus on issues that

− raise lega l uncer tainty f or provid ers and prospective provider s to inno vat e, exper iment with AI in the Uni on and

− contr ibut e to evidence-based regulator y lear ning. The super vision of the AI systems in the AI regulator y sandbo x

− should theref ore co ver their development, training, testing and validation bef ore the syste ms are placed on the

− mark et or put into ser vice, as well as the notion and occur rence of substantial modifi cation that ma y require a new

− conf or mity assessment procedure. Any signifi cant r isks identified dur ing the development and te sting of suc h AI

− syste ms should result in adequat e mitiga tion and, f ailing that, in the suspension of the development and te sting

− process. Where appropr iat e, national com petent author ities establishing AI regulator y sandbo xe s should cooperate

− with other relevant author ities, including those super vising the prot ection of fundamental r ights, and could allow f or

− the inv olvement of other actor s within the AI ecosystem suc h as national or European standardisation org anisations,

− notified bodies, testing and exper imentation f acilities, research and exper imentation labs, European Digital

− Inno vation Hubs and relevant stak eholder and civil society organisations. T o ensure unif or m implementation across

− the Uni on and economies of scale, it is appropr iate to establish common r ules f or the AI regulator y sandbo xe s ’

− im plementation and a framework f or cooperation between the relevant author ities in v olved in the super vision of the

− sandbo xe s. AI regulatory sandbo xe s established under this Regulation should be without prejudice to other la w

− allo wing f or the establishment of other sandbo xes aiming to ensure comp liance with law other than this Regulation.

− Where appropr iate, relevant comp etent author ities in ch arge of those other regulato r y sandbo xes should consider

− the benefits of using those sandbo xe s also f or the pur pose of ensur ing comp liance of AI syste ms with this

− Regulation. Upon agreement between the national compet ent author ities and the par ticipants in the AI regulator y

− sandbo x, te sting in real world conditions ma y also be operated and super vised in the framework of the AI regulator y

− sandbo x.

− (140) This Regulation should provid e the lega l basis f or the provid ers and prospective provid ers in the AI regulator y

− sandbo x to use personal data collect ed f or other pur poses f or developing cer tain AI systems in the public interest

− within the AI regulator y sandbo x, only under specif ied conditions, in accordance with Ar ticle 6(4) and Ar ticle 9(2),

− point (g), of Regulation (EU) 2016/679, and Ar ticles 5, 6 and 10 of Regulation (EU) 2018/1725, and without

− prejudice to Ar ticle 4(2) and Ar ticle 10 of Directive (EU) 2016/680. All other obliga tions of data controllers and

− r ights of data subjects under Regulations (EU) 2016/679 and (EU) 2018/1725 and Directive (EU) 2016/680 remain

− applicable. In par ticular , this Regulation should not pro vide a lega l basis in the meaning of Ar ticle 22(2), point (b) of

− Regulation (EU) 2016/679 and Ar ticle 24(2), point (b) of Regulation (EU) 2018/1725. Providers and prospective

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 35/144

− pro viders in the AI regulator y sandbo x should ensure appropr iate saf eguards and cooperate with the compet ent

− author ities, including by f ollowing their guidance and acting expeditiously and in good f aith to adequat ely mitigat e

− an y identifie d signifi cant r isks to saf ety , health, and fundamental r ights that ma y ar ise dur ing the development,

− t esting and exper imentation in that sandbo x.

− (141) In order to accelerate the process of development and the placing on the marke t of the high-r isk AI syste ms listed in

− an annex to this Regulation, it is impor tant that provid ers or prospective providers of suc h syste ms ma y also benefit

− from a specif ic regime f or te sting those systems in real w orld conditions, without par ticipating in an AI regulator y

− sandbo x. However , in suc h cases, taking into account the possible consequences of such testing on individuals, it

− should be ensured that appropr iate and sufficient guarantees and conditions are introduced by this Regulation f or

− pro viders or prospective provider s. Such guarantees should include, inter alia, requesting inf or med consent of

− natural persons to par ticipate in te sting in real wo rld conditions, with the ex ception of law enf orcement where the

− seeking of inf or med consent would prevent the AI syste m from being te sted. Consent of subjects to par ticipate in

− suc h te sting under this Regulation is distinct from, and without prejudice to, consent of data subjects f or the

− processing of their personal data under the relevant data protection law . It is also imp or tant to minimise the r isks

− and enable ove rsight by compet ent author ities and theref ore require prospective provid ers to hav e a real-wo rld

− t esting plan submitted to comp etent mark et sur veillance author ity , regist er the t esting in dedicated sections in the EU

− database subject to some limited ex ceptions , set limitations on the per iod f or which the te sting can be done and

− require additional safegua rds f or persons belonging to cer tain vulnerable groups, as well as a wr itten agreement

− defining the roles and responsibilities of prospective providers and deplo y ers and effe ctive overs ight by compet ent

− personnel in volved in the real w orld te sting. Fur ther more, it is appropr iate to envisag e additional saf eguards to

− ensure that the predictions, recommendations or decisions of the AI system can be eff ectively reversed and

− disreg arded and that personal data is protect ed and is deleted when the subjects hav e withdra wn their consent to

− par ticipate in the t esting without prejudice to their r ights as data subjects under the Union data protection law . As

− rega rds transfer of data, it is also appropr iate to envis age that data collect ed and processed f or the pur pose of te sting

− in real-world conditions should be transf er red to third countr ies only where appropr iat e and applicable saf eguards

− under Uni on law are im plemented, in par ticular in accordance with bases f or transf er of personal data under Uni on

− la w on data prot ection, while f or non-personal data appropr iate saf eguards are put in place in accordance with

− Uni on law , suc h as Regulations (EU) 2022/868 (

− 42

− ) and (EU) 2023/2854 (

− 43

− ) of the European Parliame nt and of the

− Council.

− (142) T o ensure that AI leads t o socially and environmentally beneficial outcomes, Member Stat es are encouraged to

− suppor t and promote research and development of AI solutions in suppor t of socially and envir onmentally

− beneficial outcomes, suc h as AI-based solutions to increase accessibility f or persons with disabilities, tackle

− socio-economic inequalities, or meet environmental targets, by allocating sufficient resources, including public and

− Uni on funding, and, where appropr iate and pro vided that the eligibility and selection cr ite r ia are fulfilled,

− consider ing in par ticular projects which pursue suc h objectives. Such projects should be based on the pr inciple of

− inte rdisciplinar y cooperation between AI developers, exper ts on inequality and non-discr imination, accessibility ,

− consumer , envir onmental, and digital r ights, as well as academics.

− (143) In order to promot e and protect innovation, it is impor tant that the intere sts of SMEs, including star t-ups, that are

− pro viders or deplo y ers of AI systems are take n into par ticular account. T o that end, Member Stat es should develop

− initiatives, which are targe ted at those operat ors, including on a wareness raising and inf or mation communication.

− Member States should provide SMEs, including star t-ups, that hav e a regist ered off ice or a branch in the Uni on, with

− pr ior ity access to the AI regulator y sandbo xes provided that they fulf il the eligibility conditions and selection cr ite r ia

− and without precluding other providers and prospective providers to access the sandbo xe s provided the same

− conditions and cr ite r ia are fulfilled. Member Stat es should utilise existing channels and where appropr iat e, establish

− new dedicated channels f or communication with SMEs, including star t-ups, deplo yers, other innovat ors and, as

− appropr iat e, local public author ities, to suppor t SMEs throughout their development path by providing guidance

− and responding to quer ies about the imp lementation of this Regulation. Where appropr iate, these channels should

− w ork tog ether to create synergies and ensure homog eneity in their guidance to SMEs, including star t-ups, and

− deplo y ers. Ad ditionally , Member Stat es should f acilitate the par ticipation of SMEs and other relevant stakeholders in

− the standardisation development processes. Moreover , the specific inte rests and needs of provid ers that are SMEs,

− 36/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− (

− 42

− ) Regulation (EU) 2022/868 of the European Parl iament and of the Council of 30 Ma y 2022 on European data gover nance and

− amending Regulation (EU) 2018/1724 (Data Governanc e A ct) (OJ L 152, 3.6.2022, p. 1).

− (

− 43

− ) Regulation (EU) 2023/2854 of the European Parl iament and of the Council of 13 December 2023 on har monised r ules on fair

− access t o and use of data and amending Regulation (EU) 2017/2394 and Directive (EU) 2020/1828 (Data A ct) (OJ L, 2023/2854,

− 22.12.2023, ELI: http://data.europa.eu/eli/reg/2023/2854/oj).

− including star t-ups, should be take n into account when notified bodies set conf or mity assessment f ees. The

− Commission should regularly assess the cer tif ication and compliance costs f or SMEs, including star t-ups, through

− transparent consultations and should work with Member States to lower suc h costs. For exam ple, translation costs

− relate d to mandat or y documentation and communication with author ities ma y constitute a signif icant cost f or

− pro viders and other operat ors, in par ticular those of a smaller scale. Member Stat es should possibly ensure that one

− of the languag es determin ed and accep ted b y them f or relevant provider s ’ documentation and f or communication

− with operat ors is one which is broadly understood by the larg est possible number of cross-border deplo y ers. In order

− t o address the specific needs of SMEs, including star t-ups, the Commission should provid e standardised te mplat es f or

− the areas covered by this Regulation, upon request of the Board. Additiona lly , the Commission should comp lement

− Member States’ eff or ts by providing a sing le inf or mation platf or m with easy-t o-use inf or mation with regard s to this

− Regulation f or all provid ers and deplo y ers, by organising appropr iate communication cam paigns to raise a wareness

− about the obligations ar ising from this Regulation, and b y evaluating and promoting the conve rgence of best

− practices in public procurement procedures in relation to AI systems. Medium-sized ent er pr ises which until recently

− qualified as small ent er pr ises within the meaning of the Annex to Commission Recommendation 2003/361/EC (

− 44

− )

− should hav e access to those suppor t measures, as those new medium-sized enter pr ises ma y sometimes lack the leg al

− resources and training necessar y to ensure proper understanding of, and comp liance with, this Regulation.

− (144) In order to promot e and prot ect innovation, the AI-on-demand platf or m, all relevant Uni on funding programmes

− and projects, suc h as Digital Europe Programme, Hor izon Europe, im plemented b y the Commission and the Member

− Stat es at Uni on or national level should, as appropr iat e, contr ibute to the ac hievement of the objectives of this

− Regulation.

− (145) In order to minimise the r isks to imp lementation resulting from lack of knowledg e and exper tise in the marke t as

− well as to facilitat e comp liance of provid ers, in par ticular SMEs, including star t-ups, and notif ied bodies with their

− oblig ations under this Regulation, the AI-on-demand platf or m, the European Digital Inno vation Hubs and the

− t esting and exper imentation facilities established by the Commission and the Member States at Uni on or national

− level should contr ibute t o the im plementation of this Regulation. Within their respective mission and fie lds of

− com petence, the AI-on-demand platf or m, the European Digital Inno vation Hubs and the te sting and

− exper imentation F acilities are able t o provide in par ticular tec hnical and scientific suppor t to provider s and

− notified bodies.

− (146) Moreo ver , in light of the ver y small size of some operat ors and in order to ensure propor tionality regard ing costs of

− inno vation, it is appropr iate to allow microenterp r ises to fulf il one of the most costly oblig ations, namely to

− establish a quality management syste m, in a sim plif ied manner which would reduce the administrative burden and

− the costs f or those ent er pr ises without affe cting the level of prot ection and the need f or comp liance with the

− requirements f or high-r isk AI systems. The Commission should develop guidelines to specify the elements of the

− quality managem ent syste m t o be fulf illed in this sim plif ied manner b y microent er pr ises.

− (147) It is appropr iate that the Commission f acilitates, t o the extent possible, access to testing and exper imentation

− f acilities to bodies, groups or laborat or ies established or accredit ed pursuant to an y relevant Union har monisation

− legislation and which fulf il tasks in the context of conf or mity assessment of products or devices covered by that

− Uni on har monisation legislation. This is, in par ticular , the case as rega rds exper t panels, exper t laborator ies and

− refe rence laborator ies in the fie ld of medical devices pursuant to Regulations (EU) 2017/745 and (EU) 2017/746.

− (148) This Regulation should establish a gover nance framew ork that both allo ws to coordinate and suppor t the

− application of this Regulation at national level, as well as build capabilities at Uni on level and integrat e stakeholders

− in the fie ld of AI. The effective im plementation and enf orcement of this Regulation require a gove r nance framework

− that allo ws to coordinat e and build up central exper tise at Uni on level. The AI Off ice was established by Commission

− Decision (

− 45

− ) and has as its mission to develop Union exper tise and capabilities in the fie ld of AI and to contr ibute to

− the imp lementation of Uni on law on AI. Member Stat es should f acilitate the tasks of the AI Offi ce with a view to

− suppor t the development of Uni on exper tise and capabilities at Uni on level and to strengthen the functioning of the

− digital sing le market. Fur ther more, a Board composed of representatives of the Member Stat es, a scientific panel to

− inte grate the scientific community and an advisor y f or um to contr ibute stakeh older in put to the imp lementation of

− this Regulation, at Uni on and national level, should be established. The development of Uni on exper tise and

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 37/144

− (

− 44

− ) Commission Recommendation of 6 Ma y 2003 concer ning the definition of micro, small and medium-sized ent er pr ises (OJ L 124,

− 20.5.2003, p. 36).

− (

− 45

− ) Commission Decision of 24.1.2024 establishing the European Ar tificial Intellig ence Office C(2024) 390.

− capabilities should also include making use of existing resources and exper tise, in par ticular through synergies with

− str uctures built up in the cont ext of the Union level enf orcement of other law and synergies with related initiatives at

− Uni on level, suc h as the EuroHPC Joint Und er taking and the AI te sting and exper imentation facilities under the

− Digital Europe Programme.

− (149) In order to f acilitate a smooth, eff ective and har monised implementation of this Regulation a Board should be

− established. The Board should ref lect the var ious interests of the AI eco-system and be comp osed of representatives

− of the Member Stat es. The Board should be responsible f or a number of advisor y tasks , including issuing opinions,

− recommendations, advice or contr ibuting to guidance on matt ers related t o the imp lementation of this Regulation,

− including on enf orcement matt ers, tec hnical specif ications or existing standards regarding the requirements

− established in this Regulation and pro viding advice t o the Commission and the Member States and their national

− com petent author ities on specific questions related to AI. In order to give some f lexibility to Member States in the

− designation of their representatives in the Board, suc h representatives ma y be any persons belonging to public

− entities who should hav e the relevant compet ences and po wers t o facilitat e coordination at national level and

− contr ibut e to the achi evement of the Board’s tasks. The Board should establish two standing sub-groups to provide

− a platf or m f or cooperation and exc hang e among market sur veillance author ities and notifying author ities on issues

− relate d, respectively , t o marke t sur veillance and notified bodies. The standing subgroup f or marke t sur veillance

− should act as the administrative cooperation group (ADCO) f or this Regulation within the meaning of Ar ticle 30 of

− Regulation (EU) 2019/1020. In accordance with Ar ticle 33 of that Regulation, the Commission should suppor t the

− activities of the standing subgroup f or marke t sur veillance by under taking marke t evaluations or studies, in

− par ticular with a view to identifying aspects of this Regulation requir ing specific and urgent coordination among

− mark et sur veillance author ities. The Board ma y establish other standing or tem porar y sub-groups as appropr iate f or

− the pur pose of examining specific issues. The Board should also cooperate , as appropr iate, with relevant Uni on

− bodies, exper ts groups and networks active in the context of relevant Uni on law , including in par ticular those active

− under relevant Uni on law on data, digital products and ser vices.

− (150) With a view to ensur ing the inv olvement of stakeholders in the imp lementation and application of this Regulation,

− an advisor y f or um should be established to advise and pro vide tec hnical exper tise to the Board and the Commission.

− T o ensure a var ied and balanced stak eholder representation between commercial and non-commercial interest and,

− within the category of commercial interests, with rega rds to SMEs and other under takings, the advisor y f or um

− should compr ise inter alia industr y , star t-ups, SMEs, academia, civil society , including the social par tners, as well as

− the Fundamental Rights Ag ency , ENISA, the European Committe e f or Standardization (CEN), the European

− Committ ee f or Electrotec hnical Standardization (CENELEC) and the European T elecommunications Standards

− Institut e (ETSI).

− (151) T o suppor t the imp lementation and enf orcement of this Regulation, in par ticular the monitoring activities of the AI

− Office as regards g eneral-pur pose AI models, a scientific panel of independent exper ts should be established. The

− independent exper ts constituting the scientific panel should be selected on the basis of up-to-da te scientific or

− t echnical exper tise in the field of AI and should perf or m their tasks with imp ar tiality , objectivity and ensure the

− confidentiality of inf or mation and data obtained in car r ying out their tasks and activities. T o allo w the reinf orcement

− of national capacities necessar y f or the effective enf orcement of this Regulation, Member Stat es should be able to

− request suppor t from the pool of exper ts constituting the scientific panel f or their enf orcement activities.

− (152) In order to suppor t adequate enf orcement as regard s AI systems and reinf orce the capacities of the Member Stat es,

− Uni on AI te sting suppor t str uctures should be established and made a vailable to the Member Stat es.

− (153) Member States hold a ke y role in the application and enf orcement of this Regulation. In that respect, each Member

− Stat e should designate at least one notifying author ity and at least one marke t sur veillance author ity as national

− com petent author ities f or the pur pose of super vising the application and implementation of this Regulation.

− Member Stat es ma y decide to appoint any kind of public entity to per f or m the tasks of the national compet ent

− author ities within the meaning of this Regulation, in accordance with their specific national organisational

− ch aracteristics and needs. In order to increase org anisation eff iciency on the side of Member Stat es and to set a sing le

− point of contact vis-à-vis the public and other counte r par ts at Member State and Union levels, each Member State

− should designate a marke t sur veillance author ity t o act as a sing le point of contact.

− 38/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− (154) The national compet ent author ities should ex ercise their powers independently , imp ar tially and without bias, so as

− t o saf eguard the pr inciples of objectivity of their activities and tasks and to ensure the application and

− im plementation of this Regulation. The members of these author ities should refrain from any action incom patible

− with their duties and should be subject to conf identiality r ules under this Regulation.

− (155) In order to ensure that providers of high-r isk AI systems can take into account the exper ience on the use of high-r isk

− AI systems f or impro ving their syste ms and the design and development process or can take any possible cor rective

− action in a timely manner , all providers should have a post-mark et monitoring syste m in place. Where relevant,

− post-mark et monitoring should include an analysis of the inte raction with other AI systems including other devices

− and sof tware. P ost-market monito r ing should not cover sensitive operational data of deplo y ers which are la w

− enf orcement author ities. This system is also k ey to ensure that the possible r isks emerging from AI systems which

− continue t o ‘lear n’ af te r being placed on the marke t or put into ser vice can be more effi ciently and timely addressed.

− In this cont ext, providers should also be required to hav e a syste m in place to repor t to the relevant author ities any

− ser ious incidents resulting from the use of their AI systems, meaning incident or malfunctioning leading to death or

− ser ious damage to health, ser ious and ir reversible disr up tion of the managem ent and operation of cr itical

− infrastr ucture, infr ingements of obliga tions under Uni on law intended to prot ect fundamental r ights or ser ious

− damage to proper ty or the en vironment.

− (156) In order to ensure an appropr iate and effe ctive enf orcement of the requirements and obligations set out by this

− Regulation, which is Uni on har monisation legislation, the system of marke t sur veillance and comp liance of products

− established by Regulation (EU) 2019/1020 should apply in its entirety . Marke t sur veillance author ities designated

− pursuant to this Regulation should hav e all enf orcement po wers laid do wn in this Regulation and in Regulation (EU)

− 2019/1020 and should ex ercise their powers and car r y out their duties independently , impar tially and without bias.

− Although the majorit y of AI syste ms are not subject to specific requirements and obligations under this Regulation,

− mark et sur veillance author ities ma y tak e measures in relation to all AI syste ms when the y present a r isk in

− accordance with this Regulation. Due to the specif ic nature of Uni on institutions, ag encies and bodies f alling within

− the scope of this Regulation, it is appropr iate to designate the European Data Protect ion Super visor as a compet ent

− mark et sur veillance author ity f or them. This should be without prejudice to the designation of national compet ent

− author ities by the Member Stat es. Market sur veillance activities should not affect the ability of the super vised entities

− t o car r y out their tasks independently , when suc h independence is required by Uni on law .

− (157) This Regulation is without prejudice to the compet ences, tasks , powers and independence of relevant national public

− author ities or bodies which super vise the application of Union law protect ing fundamental r ights, including equality

− bodies and data prot ection author ities. Where necessar y f or their mandat e, those national public author ities or

− bodies should also hav e access to any documentation create d under this Regulation. A specific saf eguard procedure

− should be set f or ensur ing adequat e and timely enf orcement ag ainst AI systems presenting a r isk t o health, saf ety and

− fundamental r ights. The procedure f or such AI syste ms presenting a r isk should be applied to high-r isk AI syste ms

− presenting a r isk, prohibited systems which hav e been placed on the market, put into ser vice or used in violation of

− the prohibited practices laid do wn in this Regulation and AI systems which hav e been made ava ilable in violation of

− the transparency requirements laid down in this Regulation and present a r isk.

− (158) Uni on financ ial ser vices law includes inte r nal govern ance and r isk -managem ent r ules and requirements which are

− applicable to regulate d financial institutions in the course of provision of those ser vices, including when they make

− use of AI syste ms. In order to ensure coherent application and enf orcement of the oblig ations under this Regulation

− and relevant r ules and requirements of the Uni on financial ser vices leg al acts, the compet ent author ities f or the

− super vision and enf orcement of those legal acts, in par ticular compet ent author ities as defined in Regulation (EU)

− No 575/2013 of the European Pa rliament and of the Council (

− 46

− ) and Directives 2008/48/EC (

− 47

− ), 2009/138/EC (

− 48

− ),

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 39/144

− (

− 46

− ) Regulation (EU) No 575/2013 of the European Parliament and of the Council of 26 June 2013 on pr udential requirements f or credit

− institutions and in vestment fir ms and amending Regulation (EU) No 648/2012 (OJ L 176, 27.6.2013, p. 1).

− (

− 47

− ) Directive 2008/48/EC of the European Parl iament and of the Council of 23 Apr il 2008 on credit agreements f or consumers and

− repealing Council Directive 87/102/EEC (OJ L 133, 22.5.2008, p. 66).

− (

− 48

− ) Directive 2009/138/EC of the European Parl iament and of the Council of 25 No vember 2009 on the taking-up and pursuit of the

− business of Insurance and Reinsurance (Solvency II) (OJ L 335, 17.12.2009, p. 1).

− 2013/36/EU (

− 49

− ), 2014/17/EU (

− 50

− ) and (EU) 2016/97 (

− 51

− ) of the European Parliament and of the Council, should be

− designate d, within their respective comp etences, as compet ent author ities f or the pur pose of super vising the

− im plementation of this Regulation, including f or marke t sur veillance activities, as regard s AI syste ms provid ed or

− used by regulated and super vised fi nancial institutions unless Member Stat es decide t o designate another author ity to

− fulfil these marke t sur veillance tasks . Those comp etent author ities should hav e all powers under this Regulation and

− Regulation (EU) 2019/1020 to enf orce the requirements and obligations of this Regulation, including po wers to

− car r y our ex post marke t sur veillance activities that can be integrat ed, as appropr iat e, into their existing super visor y

− mec hanisms and procedures under the relevant Union fin ancial ser vices law . It is appropr iate to envisag e that, when

− acting as market sur veillance author ities under this Regulation, the national author ities responsible f or the

− super vision of credit institutions regulate d under Directive 2013/36/EU, which are par ticipating in the Sing le

− Super visor y Mec hanism established by Council Regulation (EU) No 1024/2013 (

− 52

− ), should repor t, without dela y , to

− the European Central Bank any inf or mation identified in the course of their mark et sur veillance activities that ma y

− be of potent ial interest f or the European Central Bank’s pr udential super visor y tasks as specified in that Regulation.

− T o fur ther enhance the consistency between this Regulation and the r ules applicable to credit institutions regulated

− under Directive 2013/36/EU, it is also appropr iate to inte grate some of the provid ers ’ procedural obligations in

− relation to r isk managem ent, post marke ting monitorin g and documentation into the existing obligations and

− procedures under Directive 2013/36/EU. In order t o av oid ove rlaps, limit ed deroga tions should also be en visaged in

− relation to the quality management system of provid ers and the monitori ng obliga tion placed on deplo yers of

− high-r isk AI syste ms to the exte nt that these apply to credit institutions regulated b y Directive 2013/36/EU. The

− same regime should apply to insurance and re-insurance under takings and insurance holding companies under

− Directive 2009/138/EC and the insurance inter mediar ies under Directive (EU) 2016/97 and other types of financial

− institutions subject to requirements regarding inter nal gover nance, ar rang ements or processes established pursuant

− t o the relevant Uni on financial ser vices law to ensure consistency and equal treatment in the financ ial sect or .

− (159) Each market sur veillance author ity f or high-r isk AI syste ms in the area of biometr ics, as listed in an annex t o this

− Regulation insofa r as those syste ms are used f or the pur poses of law enf orcement, migration, asylum and border

− control management, or the administration of justice and democratic processes, should hav e effe ctive invest igative

− and cor rective powers, including at least the po wer to obtain access to all personal data that are being processed and

− t o all inf or mation necessar y f or the perf or mance of its tasks . The market sur veillance author ities should be able to

− ex ercise their powers b y acting with complet e independence. Any limitations of their access to sensitive operational

− data under this Regulation should be without prejudice to the powers confe r red to them by Directive

− (EU) 2016/680. No exclusion on disclosing data t o national data prot ection author ities under this Regulation should

− aff ect the cur rent or future powers of those author ities bey ond the scope of this Regulation.

− (160) The mark et sur veillance author ities and the Commission should be able to propose joint activities, including joint

− in vestig ations, to be conducted by marke t sur veillance author ities or market sur veillance author ities jointly with the

− Commission, that hav e the aim of promoting comp liance, identifying non-compliance, raising aw areness and

− pro viding guidance in relation to this Regulation with respect t o specific categor ies of high-r isk AI syste ms that are

− f ound to present a ser ious r isk across tw o or more Member Stat es. Joint activities to promote compliance should be

− car r ied out in accordance with Ar ticle 9 of Regulation (EU) 2019/1020. The AI Office should provid e coordination

− suppor t f or joint investig ations.

− (161) It is necessar y to clar ify the responsibilities and comp et ences at Uni on and national level as regard s AI syste ms that

− are built on g eneral-pur pose AI models. T o av oid ove rlapping comp etences, where an AI syste m is based on

− a g eneral-pur pose AI model and the model and system are provid ed b y the same provider , the super vision should

− 40/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− (

− 49

− ) Directive 2013/36/EU of the European Parl iament and of the Council of 26 June 2013 on access to the activity of credit institutions

− and the pr udential super vision of credit institutions and in vestment fir ms, amending Directive 2002/87/EC and repealing

− Directives 2006/48/EC and 2006/49/EC (OJ L 176, 27.6.2013, p. 338).

− (

− 50

− ) Directive 2014/17/EU of the European Parl iament and of the Council of 4 Febr uar y 2014 on credit agreements f or consumers

− relating to residential immov able proper ty and amending Directives 2008/48/EC and 2013/36/EU and Regulation (EU)

− No 1093/2010 (OJ L 60, 28.2.2014, p. 34).

− (

− 51

− ) Directive (EU) 2016/97 of the European Parliament and of the Council of 20 Januar y 2016 on insurance distr ibution (OJ L 26,

− 2.2.2016, p. 19).

− (

− 52

− ) Council Regulation (EU) No 1024/2013 of 15 Oct ober 2013 conf er r ing specific tasks on the European Central Bank concer ning

− policies relating t o the pr udential super vision of credit institutions (OJ L 287, 29.10.2013, p. 63).

− tak e place at Union level through the AI Office, which should have the powers of a market sur veillance author ity

− within the meaning of Regulation (EU) 2019/1020 f or this pur pose. In all other cases, national market sur veillance

− author ities remain responsible f or the super vision of AI syste ms. How ever , f or g eneral-pur pose AI systems that can

− be used directly by deplo yers f or at least one pur pose that is classified as high-r isk, marke t sur veillance author ities

− should cooperate with the AI Office to car r y out evaluations of compliance and inf or m the Board and other marke t

− sur veillance author ities according ly . Fur ther more, market sur veillance author ities should be able to request

− assistance from the AI Off ice where the marke t sur veillance author ity is unable to conclude an invest igation on

− a high-r isk AI syste m because of its inability to access cer tain inf or mation related to the g eneral-pur pose AI model

− on which the high-r isk AI syste m is built. In suc h cases, the procedure regarding mutual assistance in cross-border

− cases in Chapt er VI of Regulation (EU) 2019/1020 should apply mutatis mutandis.

− (162) T o mak e best use of the centralised Uni on exper tise and synergies at Uni on level, the po wers of super vision and

− enf orcement of the obliga tions on provid ers of g eneral-pur pose AI models should be a comp etence of the

− Commission. The AI Office should be able to car r y out all necessar y actions t o monitor the effe ctive im plementation

− of this Regulation as rega rds general-pur pose AI models. It should be able to in vestigat e possible infr ingements of

− the r ules on providers of general-pur pose AI models both on its ow n initiative, f ollowing the results of its

− monito r ing activities, or upon request from market sur veillance author ities in line with the conditions set out in this

− Regulation. T o suppor t effe ctive monitoring of the AI Offi ce, it should provide f or the possibility that downstream

− pro viders lodg e complaints about possible infr ing ements of the r ules on provid ers of g eneral-pur pose AI models and

− syste ms.

− (163) With a view to comp lementing the governance systems f or general-pur pose AI models, the scientific panel should

− suppor t the monitoring activities of the AI Offi ce and ma y , in cer tain cases, provide qualified aler ts to the AI Off ice

− which tr igger f ollow-up s, suc h as invest igations. This should be the case where the scientific panel has reason to

− suspect that a general-purpo se AI model poses a concrete and identifi able r isk at Union level. Fur ther more, this

− should be the case where the scientific panel has reason to suspect that a g eneral-pur pose AI model meets the cr ite r ia

− that wo uld lead to a classif ication as g eneral-pur pose AI model with systemic r isk. T o equip the scientific panel with

− the inf or mation necessar y f or the perf or mance of those tasks , there should be a mec hanism whereby the scientific

− panel can request the Commission to require documentation or inf or mation from a provid er .

− (164) The AI Offi ce should be able t o take the necessar y actions to monitor the eff ective imp lementation of and

− com pliance with the obliga tions f or provid ers of g eneral-pur pose AI models laid do wn in this Regulation. The AI

− Office should be able to invest igat e possible infr ingements in accordance with the powers provided f or in this

− Regulation, including by requesting documentation and inf or mation, b y conducting evaluations, as well as by

− requesting measures from provid ers of g eneral-pur pose AI models. When conducting evaluations, in order to make

− use of independent exper tise, the AI Off ice should be able to in v olve independent exper ts to car r y out the

− evaluations on its behalf. Compliance with the obliga tions should be enf orceable, inter alia, through requests to take

− appropr iat e measures, including r isk mitiga tion measures in the case of identifie d syste mic r isks as well as restr icting

− the making availa ble on the market, withdra wing or recalling the model. As a saf eguard, where needed bey ond the

− procedural r ights provided f or in this Regulation, provid ers of g eneral-pur pose AI models should hav e the

− procedural r ights provid ed f or in Ar ticle 18 of Regulation (EU) 2019/1020, which should apply mutatis mutandis,

− without prejudice to more specif ic procedural r ights provid ed f or by this Regulation.

− (165) The development of AI syste ms other than high-r isk AI systems in accordance with the requirements of this

− Regulation ma y lead to a larg er uptak e of ethical and tr ustw or th y AI in the Uni on. Providers of AI syste ms that are

− not high-r isk should be encouraged to create codes of conduct, including related gover nance mechanisms, intended

− t o f oste r the v oluntar y application of some or all of the mandator y requirements applicable to high-r isk AI systems,

− adap te d in light of the intended pur pose of the systems and the low er r isk in volved and taking into account the

− a vailable te ch nical solutions and industr y best practices suc h as model and data cards. Provi ders and, as appropr iate,

− deplo y ers of all AI systems, high-r isk or not, and AI models should also be encourage d to apply on a v oluntar y basis

− additional requirements relate d, f or exam ple, to the elements of the Uni on’s Ethics Guidelines f or T r ustw or th y AI,

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 41/144

− en vironmental sustainability , AI literac y measures, inclusive and diverse design and development of AI systems,

− including atte ntion to vulnerable persons and accessibility to persons with disability , stakeholders’ par ticipation with

− the in volvement, as appropr iate, of relevant stak eholders such as business and civil society organisations, academia,

− researc h org anisations, trade unions and consumer prot ection org anisations in the design and development of AI

− syste ms, and diversity of the development t eams, including g ender balance. T o ensure that the v oluntar y codes of

− conduct are effe ctive, they should be based on clear objectives and ke y perf or mance indicators to measure the

− ac hievement of those objectives. They should also be developed in an inclusive wa y , as appropr iate, with the

− in v olvement of relevant stakeholders suc h as business and civil society organisations, academia, research

− org anisations, trade unions and consumer prot ection organisation. The Commission ma y develop initiatives,

− including of a sect oral nature, to f acilitate the lowering of tec hnical bar r iers hinder ing cross-border exc hang e of data

− f or AI development, including on data access infrastr ucture, semantic and tec hnical interoperability of different types

− of data.

− (166) It is im por tant that AI systems relate d to products that are not high-r isk in accordance with this Regulation and thus

− are not required to comp ly with the requirements set out f or high-r isk AI systems are never theless safe when placed

− on the marke t or put into ser vice. T o contr ibut e to this objective, Regulation (EU) 2023/988 of the European

− P arliament and of the Council (

− 53

− ) would apply as a saf ety net.

− (167) In order to ensure tr ustful and constr uctive cooperation of comp et ent author ities on Uni on and national level, all

− par ties inv olved in the application of this Regulation should respect the conf identiality of inf or mation and data

− obtained in car r ying out their tasks , in accordance with Uni on or national law . They should car r y out their tasks and

− activities in suc h a manner as to prot ect, in par ticular , intellectual proper ty r ights, confi dential business inf or mation

− and trade secrets, the effective im plementation of this Regulation, public and national secur ity interests, the integr ity

− of cr iminal and administrative proceedings, and the integr ity of classif ied inf or mation.

− (168) Comp liance with this Regulation should be enf orceable by means of the imp osition of penalties and other

− enf orcement measures. Member States should take all necessar y measures to ensure that the provisions of this

− Regulation are implement ed, including b y la ying do wn effe ctive, propor tionate and dissuasive penalties f or their

− infr ing ement, and to respect the ne bis in idem pr inciple. In order to strengthen and har monise administrative

− penalties f or infr ingem ent of this Regulation, the upper limits f or setting the administrative fine s f or cer tain specific

− infr ing ements should be laid do wn. When assessing the amount of the fine s, Member States should, in each

− individual case, take into account all relevant circumstances of the specific situation, with due regard in par ticular to

− the nature, gra vity and duration of the infr ing ement and of its consequences and to the size of the provid er , in

− par ticular if the provider is an SME, including a star t-up. The European Data Protection Super visor should hav e the

− po wer to im pose fines on Uni on institutions, agencies and bodies falling within the scope of this Regulation.

− (169) Comp liance with the obligations on pro viders of g eneral-pur pose AI models imposed under this Regulation should

− be enf orceable, inte r alia, b y means of fines. T o that end, appropr iate levels of fi nes should also be laid do wn f or

− infr ing ement of those obligati ons, including the f ailure t o comply with measures request ed by the Commission in

− accordance with this Regulation, subject to appropr iate limitation per iods in accordance with the pr inciple of

− propor tionality . All decisions taken by the Commission under this Regulation are subject to review by the Cour t of

− Justice of the European Uni on in accordance with the TFEU, including the unlimit ed jur isdiction of the Cour t of

− Justice with rega rd to penalties pursuant to Ar ticle 261 TFEU.

− (170) Uni on and national law already provide eff ective remedies t o natural and leg al persons whose r ights and freedoms

− are adversely aff ected by the use of AI syste ms. Without prejudice to those remedies, an y natural or lega l person that

− has grounds to consider that there has been an infr ingem ent of this Regulation should be entitled to lodg e

− a complaint to the relevant mark et sur veillance author ity .

− (171) Aff ecte d persons should hav e the r ight to obtain an explanation where a deplo y er ’s decision is based mainly upon

− the output from cer tain high-r isk AI syste ms that f all within the scope of this Regulation and where that decision

− produces leg al effe cts or similarly significantly affe cts those persons in a wa y that the y consider to have an adverse

− 42/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− (

− 53

− ) Regulation (EU) 2023/988 of the European Parliament and of the Council of 10 Ma y 2023 on g eneral product safe ty , amending

− Regulation (EU) No 1025/2012 of the European Parliament and of the Council and Directive (EU) 2020/1828 of the European

− Parl iament and the Council, and repealing Directive 2001/95/EC of the European Parl iament and of the Council and Council

− Directive 87/357/EEC (OJ L 135, 23.5.2023, p. 1).

− im pact on their health, safety or fundamental r ights. That explanation should be clear and meaningful and should

− pro vide a basis on which the affe cted persons are able to exercise their r ights. The r ight to obtain an explanation

− should not apply to the use of AI syste ms f or which ex ceptions or restr ictions f ollow from Uni on or national law

− and should apply only to the extent this r ight is not already provid ed f or under Uni on law .

− (172) P ersons acting as whistleblo wers on the infr ing ements of this Regulation should be prot ected under the Union law .

− Directive (EU) 2019/1937 of the European Pa rliament and of the Council (

− 54

− ) should theref ore apply to the repor ting

− of infr ingements of this Regulation and the protect ion of persons repor ting suc h infr ingements.

− (173) In order to ensure that the regulato r y framework can be adap te d where necessar y , the power t o adopt acts in

− accordance with Ar ticle 290 TFEU should be deleg ated to the Commission t o amend the conditions under which an

− AI system is not to be considered to be high-r isk, the list of high-r isk AI syste ms, the provisions regarding te ch nical

− documentation, the cont ent of the EU declaration of conf or mity the provisions regard ing the conf or mity assessment

− procedures, the provisions establishing the high-r isk AI syste ms t o which the conf or mity assessment procedure

− based on assessment of the quality management syste m and assessment of the te ch nical documentation should

− apply , the threshold, bench mark s and indicators, including by supplementing those bench mark s and indicator s, in

− the r ules f or the classification of general-pur pose AI models with syste mic r isk, the cr ite r ia f or the designation of

− g eneral-pur pose AI models with syste mic r isk, the te ch nical documentation f or providers of g eneral-pur pose AI

− models and the transparency inf or mation f or providers of g eneral-pur pose AI models. It is of par ticular imp or tance

− that the Commission car r y out appropr iate consultations dur ing its preparatory wo rk , including at exper t level, and

− that those consultations be conducte d in accordance with the pr inciples laid down in the Inter institutional

− Agreement of 13 Apr il 2016 on Better Law -Making (

− 55

− ). In par ticular , t o ensure equal par ticipation in the

− preparation of delegat ed acts, the European P arliament and the Council receive all documents at the same time as

− Member Stat es’ exper ts, and their exper ts syste matically hav e access to meetings of Commission exper t groups

− dealing with the preparation of delegat ed acts.

− (174) Given the rapid tec hnological developments and the t echnical exper tise required to effe ctively apply this Regulation,

− the Commission should evaluate and review this Regulation by 2 Augu st 2029 and ever y f our y ears thereaf te r and

− repor t to the European Parliament and the Council. In addition, taking into account the implications f or the scope of

− this Regulation, the Commission should car r y out an assessment of the need to amend the list of high-r isk AI

− syste ms and the list of prohibited practices once a y ear . Moreove r , by 2 August 2028 and ever y f our y ears thereaf te r ,

− the Commission should evaluate and repor t t o the European Parliament and to the Council on the need to amend

− the list of high-r isk areas headings in the annex to this Regulation, the AI systems within the scope of the

− transparency obligations, the effectiveness of the super vision and gover nance syste m and the progress on the

− development of standardisation deliverables on energy eff icient development of g eneral-pur pose AI models,

− including the need f or fur ther measures or actions. Finally , by 2 Au gust 2028 and ever y three y ears thereaf te r , the

− Commission should evaluate the im pact and effectiveness of v oluntar y codes of conduct to f oste r the application of

− the requirements provided f or high-r isk AI systems in the case of AI systems other than high-r isk AI syste ms and

− possibly other additional requirements f or suc h AI syste ms.

− (175) In order to ensure unif or m conditions f or the imp lementation of this Regulation, impl ementing po wers should be

− conf er red on the Commission. Those powers should be ex ercised in accordance with Regulation (EU) No 182/2011

− of the European Pa rliament and of the Council (

− 56

− ).

− (176) Since the objective of this Regulation, namely to imp rove the functioning of the inte r nal mark et and to promot e the

− up take of human centr ic and tr ustw or th y AI, while ensur ing a high level of protection of health, saf ety , fundamental

− r ights enshr ined in the Char te r , including democracy , the r ule of law and environmental prot ection aga inst har mful

− eff ects of AI syste ms in the Uni on and suppor ting inno vation, cannot be sufficiently achieved by the Member Stat es

− and can rather , by reason of the scale or effe cts of the action, be bett er ac hieved at Uni on level, the Uni on ma y adop t

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 43/144

− (

− 54

− ) Directive (EU) 2019/1937 of the European Parliament and of the Council of 23 Oct ober 2019 on the protection of persons who

− repor t breaches of Union la w (OJ L 305, 26.11.2019, p. 17).

− (

− 55

− ) OJ L 123, 12.5.2016, p. 1.

− (

− 56

− ) Regulation (EU) No 182/2011 of the European Parliament and of the Council of 16 Febr uar y 2011 la ying down the r ules and

− g eneral pr inciples concer ning mechanisms f or control b y Member States of the Commission’s exerci se of im plementing po wers (OJ

− L 55, 28.2.2011, p. 13).

− measures in accordance with the pr inciple of subsidiar ity as set out in Ar ticle 5 TEU. In accordance with the

− pr inciple of propor tionality as set out in that Ar ticle, this Regulation does not go bey ond what is necessar y in order

− t o achieve that objective.

− (177) In order to ensure lega l cer tainty , ensure an appropr iat e adaptation per iod f or operat ors and av oid disr uption to the

− mark et, including b y ensur ing continuity of the use of AI syste ms, it is appropr iate that this Regulation applies to the

− high-r isk AI syste ms that hav e been placed on the marke t or put into ser vice bef ore the g eneral date of application

− thereof, only if, from that date, those systems are subject to signifi cant chang es in their design or intende d pur pose.

− It is appropr iate to clar ify that, in this respect, the concept of significant chang e should be understood as equivalent

− in substance to the notion of substantial modif ication, which is used with regard only to high-r isk AI syste ms

− pursuant t o this Regulation. On an ex ceptiona l basis and in light of public accountability , operat ors of AI syste ms

− which are comp onents of the larg e-scale IT systems established b y the lega l acts listed in an annex to this Regulation

− and operators of high-r isk AI syste ms that are intended to be used by public author ities should, respectively , take the

− necessar y st eps to comp ly with the requirements of this Regulation b y end of 2030 and by 2 August 2030.

− (178) Provi ders of high-r isk AI systems are encourage d to star t to comply , on a v oluntar y basis, with the relevant

− oblig ations of this Regulation already dur ing the transitional per iod.

− (179) This Regulation should apply from 2 August 2026. How ever , taking into account the unaccepta ble r isk associated

− with the use of AI in cer tain wa ys, the prohibitions as well as the g eneral provis ions of this Regulation should already

− apply from 2 Febr uar y 2025. While the full effe ct of those prohibitions f ollows with the establishment of the

− gove r nance and enf orcement of this Regulation, anticipating the application of the prohibitions is imp or tant to take

− account of unaccepta ble r isks and t o hav e an effect on other procedures, suc h as in civil law . Moreove r , the

− infrastr ucture related to the gove r nance and the conf or mity assessment syste m should be operational bef ore

− 2 August 2026, theref ore the provisions on notified bodies and governance str ucture should apply from 2 August

− 2025. Given the rapid pace of t echnological advancements and adop tion of general-pur pose AI models, obligations

− f or provid ers of g eneral-pur pose AI models should apply from 2 Au gust 2025. Codes of practice should be ready by

− 2 Ma y 2025 in view of enabling provid ers to demonstrate com pliance on time. The AI Office should ensure that

− classification r ules and procedures are up to date in light of te chnologi cal developments. In addition, Member Stat es

− should la y down and notify to the Commission the r ules on penalties, including administrative fines, and ensure that

− the y are properly and effe ctively im plemented by the date of application of this Regulation. Theref ore the provis ions

− on penalties should apply from 2 Au gust 2025.

− (180) The European Data Protection Super visor and the European Data Prot ection Board were consult ed in accordance

− with Ar ticle 42(1) and (2) of Regulation (EU) 2018/1725 and delivered their joint opinion on 18 June 2021,

− HA VE ADOPTED THIS REGUL A TION:

− CHAPTER I

− GENERAL PR O VISIONS

− Ar ticle 1

− Subject matter`

− 1. The pur pose of this Regulation is to impro ve the functioning of the inter nal marke t and promote the up take of

− human-centr ic and tr ustw or thy ar tif icial intellig ence (AI), while ensur ing a high level of prot ection of health, saf ety ,

− fundamental r ights enshr ined in the Char te r , including democracy , the r ule of la w and environmental protection, aga inst

− the har mful eff ects of AI systems in the Uni on and suppor ting inno vation.

− 2. This Regulation la ys do wn:

− (a) har monised r ules f or the placing on the marke t, the putting into ser vice, and the use of AI systems in the Uni on;

− 44/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− (b) prohibitions of cer tain AI practices;

− (c) specific requirements f or high-r isk AI systems and oblig ations f or operators of suc h systems;

− (d) har monised transparency r ules f or cer tain AI systems;

− (e) har monised r ules f or the placing on the marke t of general-pur pose AI models;

− (f) r ules on mark et monitoring, marke t sur veillance, gove r nance and enf orcement ;

− (g) measures to suppor t innovation, with a par ticular f ocus on SMEs, including star t-ups.

− Ar ticle 2

− Scope

− 1. This Regulation applies to:

− (a) provid ers placing on the marke t or putting into ser vice AI syste ms or placing on the market general-pur pose AI models

− in the Uni on, ir respective of whether those pro viders are established or locat ed within the Union or in a third countr y ;

− (b) deplo yers of AI syste ms that have their place of establishment or are located within the Uni on;

− (c) provid ers and deplo y ers of AI syste ms that hav e their place of establishment or are locate d in a third countr y , where the

− output produced by the AI system is used in the Union;

− (d) imp or te rs and distr ibut ors of AI systems;

− (e) product manufacturers placing on the marke t or putting into ser vice an AI system tog ether with their product and

− under their o wn name or trademark;

− (f) author ised representatives of provider s, which are not established in the Uni on;

− (g) affe cted persons that are located in the Union.

− 2. For AI syste ms classif ied as high-r isk AI systems in accordance with Ar ticle 6(1) related to products covered by the

− Uni on har monisation legislation listed in Section B of Annex I, only Ar ticle 6(1), Ar ticles 102 to 109 and Ar ticle 112 apply .

− Ar ticle 57 applies only in so f ar as the requirements f or high-r isk AI systems under this Regulation have been integrat ed in

− that Union har monisation legislation.

− 3. This Regulation does not apply to areas outside the scope of Union law , and shall not, in any event, affe ct the

− compet ences of the Member States concer ning national secur ity , regardless of the type of entity entr uste d by the Member

− Stat es with car r ying out tasks in relation to those comp etences.

− This Regulation does not apply to AI syste ms where and in so far they are placed on the marke t, put into ser vice, or used

− with or without modif ication exclusively f or militar y , defe nce or national secur ity pur poses, rega rdless of the type of entity

− car r ying out those activities.

− This Regulation does not apply t o AI systems which are not placed on the market or put into ser vice in the Union, where

− the output is used in the Uni on ex clusively f or militar y , defe nce or national secur ity pur poses, regardless of the type of

− entity car r ying out those activities.

− 4. This Regulation applies neither to public author ities in a third countr y nor to inte r national org anisations f alling

− within the scope of this Regulation pursuant to paragraph 1, where those author ities or organisations use AI syste ms in the

− framew ork of international cooperation or agreements f or law enf orcement and judicial cooperation with the Union or

− with one or more Member Stat es, provided that suc h a third countr y or intern ational org anisation provides adequate

− saf eguards with respect t o the prot ection of fundamental r ights and freedoms of individuals.

− 5. This Regulation shall not affect the application of the provisions on the liability of provider s of interm ediar y ser vices

− as set out in Chapt er II of Regulation (EU) 2022/2065.

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 45/144

− 6. This Regulation does not apply to AI systems or AI models, including their output, specif ically developed and put into

− ser vice f or the sole pur pose of scientifi c researc h and development.

− 7. Union law on the prot ection of personal data, pr ivacy and the conf identiality of communications applies to personal

− data processed in connection with the r ights and obliga tions laid down in this Regulation. This Regulation shall not affect

− Regulation (EU) 2016/679 or (EU) 2018/1725, or Directive 2002/58/EC or (EU) 2016/680, without prejudice to Ar ticle

− 10(5) and Ar ticle 59 of this Regulation.

− 8. This Regulation does not apply to any researc h, te sting or development activity regard ing AI syste ms or AI models

− pr ior to their being placed on the marke t or put into ser vice. Such activities shall be conducted in accordance with

− applicable Union law . T esting in real world conditions shall not be co vered by that ex clusion.

− 9. This Regulation is without prejudice t o the r ules laid do wn by other Uni on leg al acts related to consumer prot ection

− and product safety .

− 10. This Regulation does not apply t o obligations of deplo y ers who are natural persons using AI syste ms in the course of

− a purely personal non-professional activity .

− 11. This Regulation does not preclude the Uni on or Member States from maintaining or introducing la ws, regulations or

− administrative provis ions which are more fa vourable to w orkers in te r ms of protect ing their r ights in respect of the use of

− AI systems by em plo y ers, or from encouraging or allowi ng the application of collective agreements which are more

− f av ourable to wo rkers.

− 12. This Regulation does not apply to AI systems released under free and open-source licences, unless they are placed on

− the mark et or put into ser vice as high-r isk AI systems or as an AI system that f alls under Ar ticle 5 or 50.

− Ar ticle 3

− Def initions

− For the pur poses of this Regulation, the f ollowing def initions apply:

− (1) ‘ AI syste m’ means a machine-based system that is designed to operat e with var ying levels of autonom y and that ma y

− exhibit adap tiveness af te r deplo yment, and that, f or explicit or implicit objectives, inf ers, from the input it receives,

− ho w to g enerate outputs such as predictions, content, recommendations, or decisions that can inf luence phys ical or

− vir tual environments;

− (2) ‘r isk’ means the combination of the probability of an occur rence of har m and the sever ity of that har m;

− (3) ‘provi der ’ means a natural or lega l person, public author ity , age ncy or other body that develops an AI system or

− a general-pur pose AI model or that has an AI syste m or a g eneral-pur pose AI model developed and places it on the

− mark et or puts the AI syste m into ser vice under its own name or trademark, whether f or pa yment or free of charg e;

− (4) ‘deplo yer ’ means a natural or lega l person, public author ity , age ncy or other body using an AI system under its

− author ity ex cept where the AI system is used in the course of a personal non-professional activity ;

− (5) ‘author ised representative ’ means a natural or leg al person located or established in the Uni on who has received and

− accep te d a wr itten mandate from a provider of an AI system or a g eneral-pur pose AI model to, respectively , perf or m

− and car r y out on its behalf the obliga tions and procedures established by this Regulation;

− (6) ‘imp or ter ’ means a natural or lega l person locate d or established in the Uni on that places on the market an AI system

− that bears the name or trademark of a natural or lega l person established in a third countr y ;

− (7) ‘distr ibutor ’ means a natural or leg al person in the supply chain, other than the pro vider or the im por ter , that mak es

− an AI syste m available on the Union marke t ;

− (8) ‘operat or ’ means a provid er , product manufa cturer , deplo y er , author ised representative, imp or ter or distr ibut or;

− 46/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− (9) ‘placing on the marke t ’ means the fir st making available of an AI system or a general-purpo se AI model on the Uni on

− mark et ;

− (10) ‘making ava ilable on the market’ means the supply of an AI syste m or a g eneral-pur pose AI model f or distr ibution or

− use on the Uni on market in the course of a commercial activity , whether in retur n f or pa yment or free of ch arg e;

− (11) ‘putting into ser vice ’ means the supply of an AI system f or first use directly to the deplo y er or f or ow n use in the Uni on

− f or its intended pur pose;

− (12) ‘int ended pur pose’ means the use f or which an AI syste m is intended by the provider , including the specif ic cont ext

− and conditions of use, as specif ied in the inf or mation supplied by the provid er in the instr uctions f or use, promotional

− or sales materials and stat ements, as well as in the tec hnical documentation;

− (13) ‘reasonably f oreseeable misuse’ means the use of an AI syste m in a wa y that is not in accordance with its intended

− pur pose, but which ma y result from reasonably f oreseeable human behavi our or interaction with other systems,

− including other AI systems;

− (14) ‘saf ety component’ means a component of a product or of an AI system which fulfils a safety function f or that product

− or AI system, or the f ailure or malfunctioning of which endang ers the health and safety of persons or proper ty ;

− (15) ‘instr uctions f or use ’ means the inf or mation provided by the provider t o inf or m the deplo y er of, in par ticular , an AI

− syste m’s intended pur pose and proper use;

− (16) ‘recall of an AI syste m’ means an y measure aiming to ac hieve the retur n to the provider or taking out of ser vice or

− disabling the use of an AI system made a vailable to deplo yers;

− (17) ‘withdra wal of an AI system’ means any measure aiming to prevent an AI system in the supply ch ain being made

− a vailable on the mark et ;

− (18) ‘perf or mance of an AI syste m’ means the ability of an AI syste m t o achieve its intended pur pose;

− (19) ‘notifying author ity’ means the national author ity responsible f or setting up and car r ying out the necessar y procedures

− f or the assessment, designation and notification of conf or mity assessment bodies and f or their monito r ing;

− (20) ‘conf or mity assessment’ means the process of demonstrating whether the requirements set out in Chapt er III, Section 2

− relating to a high-r isk AI system have been fulfilled;

− (21) ‘conf or mity assessment body’ means a body that perfo r ms third-par ty conf or mity assessment activities, including

− t esting, cer tif ication and inspection;

− (22) ‘notified body’ means a conf or mity assessment body notif ied in accordance with this Regulation and other relevant

− Uni on har monisation legislation;

− (23) ‘substantial modifi cation’ means a ch ange to an AI syste m af te r its placing on the marke t or putting into ser vice which

− is not f oreseen or planned in the initial conf or mity assessment car r ied out b y the provid er and as a result of which the

− com pliance of the AI system with the requirements set out in Chap ter III, Section 2 is affe cted or results in

− a modif ication to the inte nded pur pose f or which the AI syste m has been assessed;

− (24) ‘CE marking’ means a marking b y which a provid er indicates that an AI system is in conf or mity with the requirements

− set out in Chapt er III, Section 2 and other applicable Uni on har monisation legislation pro viding f or its affixing;

− (25) ‘post-mark et monitoring syste m’ means all activities car r ied out by provid ers of AI systems to collect and review

− exper ience gained from the use of AI systems they place on the marke t or put into ser vice f or the pur pose of

− identifying any need to immediate ly apply any necessar y cor rective or preventive actions;

− (26) ‘marke t sur veillance author ity’ means the national author ity car r ying out the activities and taking the measures

− pursuant to Regulation (EU) 2019/1020;

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 47/144

− (27) ‘har monised standard’ means a har monised standard as def ined in Ar ticle 2(1), point (c), of Regulation (EU)

− No 1025/2012;

− (28) ‘common specif ication’ means a set of te chnical specif ications as defined in Ar ticle 2, point (4) of Regulation (EU)

− No 1025/2012, providing means to compl y with cer tain requirements established under this Regulation;

− (29) ‘training data’ means data used f or training an AI system through fi tting its lear nable paramet ers;

− (30) ‘validation data’ means data used f or providing an evaluation of the trained AI system and f or tuning its non-lear nable

− paramet ers and its lear ning process in order , inter alia, to prevent under fit ting or ove r fitting;

− (31) ‘validation data set’ means a separate data set or par t of the training data set, either as a fixed or var iable split ;

− (32) ‘te sting data’ means data used f or providing an independent evaluation of the AI system in order t o conf ir m the

− expect ed perform ance of that syste m bef ore its placing on the market or putting into ser vice;

− (33) ‘in put data’ means data provided to or directly acquired by an AI syste m on the basis of which the syste m produces an

− output ;

− (34) ‘biometr ic data’ means personal data resulting from specific te chnical processing relating t o the physical , physiolo gical

− or behavi oural ch aracter istics of a natural person, such as facial images or dactyloscopic data;

− (35) ‘biometr ic identifica tion’ means the automat ed recognition of physical, physiolo gical, behavioural, or psy chological

− human features f or the pur pose of establishing the identity of a natural person b y comp ar ing biometr ic data of that

− individual to biometr ic data of individuals st ored in a database;

− (36) ‘biometr ic ver ificati on’ means the automa ted, one-to-one ver ification, including authentication, of the identity of

− natural persons by compari ng their biometr ic data to previously provided biometr ic data;

− (37) ‘special categori es of personal data’ means the categori es of personal data refer red to in Ar ticle 9(1) of Regulation (EU)

− 2016/679, Ar ticle 10 of Directive (EU) 2016/680 and Ar ticle 10(1) of Regulation (EU) 2018/1725;

− (38) ‘sensitive operational data’ means operational data related t o activities of prevention, detection, in vestig ation or

− prosecution of cr iminal offe nces, the disclosure of which could jeopardise the inte gr ity of cr iminal proceedings;

− (39) ‘emotion recognition syste m’ means an AI syste m f or the pur pose of identifying or inferr ing emotions or inte ntions of

− natural persons on the basis of their biometr ic data;

− (40) ‘biometr ic categor isation syste m’ means an AI system f or the pur pose of assigning natural persons to specif ic

− cate gor ies on the basis of their biometr ic data, unless it is ancillar y to another commercial ser vice and str ictly

− necessar y f or objective te chnical reasons;

− (41) ‘remot e biometr ic identification system’ means an AI system f or the pur pose of identifying natural persons, without

− their active in volvement, typically at a distance through the comp ar ison of a person’s biometr ic data with the

− biometr ic data contained in a refere nce database;

− (42) ‘real-time remote biometr ic identifica tion syste m’ means a remot e biometr ic identification system, whereby the

− capt ur ing of biometr ic data, the comp ar ison and the identification all occur without a signifi cant dela y , compr ising

− not only instant identifica tion, but also limited shor t dela ys in order to av oid circum vention;

− (43) ‘post-remot e biometr ic identifica tion syste m’ means a remot e biometr ic identification system other than a real-time

− remot e biometr ic identification system;

− (44) ‘publicly accessible space ’ means any publicly or pr ivately owned physica l place accessible t o an undet er mined number

− of natural persons, regardless of whether cer tain conditions f or access ma y apply , and regard less of the pote ntial

− capacity restr ictions;

− 48/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− (45) ‘law enf orcement author ity’ means:

− (a) an y public author ity comp etent f or the prevention, invest igation, detection or prosecution of cr iminal offences or

− the ex ecution of cr iminal penalties, including the saf eguarding aga inst and the prevention of threats to public

− secur ity ; or

− (b) an y other body or entity entr uste d by Member Stat e law to exercise public author ity and public powers f or the

− pur poses of the prevention, investig ation, detect ion or prosecution of cr iminal offe nces or the execution of

− cr iminal penalties, including the safeguarding against and the prevention of threats to public secur ity ;

− (46) ‘law enf orcement’ means activities car r ied out by law enf orcement author ities or on their behalf f or the prevention,

− in vestig ation, detection or prosecution of cr iminal offe nces or the execution of cr iminal penalties, including

− saf eguarding against and preventing threats to public secur ity ;

− (47) ‘ AI Off ice’ means the Commission’s function of contr ibuting to the im plementation, monitorin g and super vision of AI

− syste ms and general-pur pose AI models, and AI gover nance, provid ed f or in Commission Decision of 24 Januar y

− 2024; refe rences in this Regulation to the AI Office shall be constr ued as refe rences to the Commission;

− (48) ‘national comp et ent author ity’ means a notifying author ity or a marke t sur veillance author ity ; as regard s AI syste ms

− put into ser vice or used by Union institutions, agencies, off ices and bodies, refe rences to national compet ent

− author ities or market sur veillance author ities in this Regulation shall be constr ued as refe rences to the European Data

− Prot ection Super visor ;

− (49) ‘ser ious incident ’ means an incident or malfunctioning of an AI syste m that directly or indirectly leads to any of the

− f ollowing:

− (a) the death of a person, or ser ious har m to a person’s health;

− (b) a ser ious and ir reversible disr uption of the managem ent or operation of cr itical infrastr ucture;

− (c) the infr ingement of obliga tions under Uni on law intended to protect fundamental r ights;

− (d) ser ious har m to proper ty or the environment ;

− (50) ‘personal data’ means personal data as defined in Ar ticle 4, point (1), of Regulation (EU) 2016/679;

− (51) ‘non-personal data’ means data other than personal data as defined in Ar ticle 4, point (1), of Regulation (EU)

− 2016/679;

− (52) ‘profiling’ means prof iling as def ined in Ar ticle 4, point (4), of Regulation (EU) 2016/679;

− (53) ‘real-w orld te sting plan’ means a document that descr ibes the objectives, methodology , g eographical, population and

− t emporal scope, monitorin g, organisation and conduct of t esting in real-world conditions;

− (54) ‘sandbo x plan’ means a document agreed between the par ticipating provid er and the comp etent author ity descr ibing

− the objectives, conditions, timeframe, methodology and requirements f or the activities car r ied out within the sandbo x;

− (55) ‘ AI regulato r y sandbo x’ means a controlled framework set up by a comp etent author ity which offers provid ers or

− prospective provider s of AI syste ms the possibility t o develop, train, validate and te st, where appropr iat e in real-wo rld

− conditions, an innovative AI system, pursuant to a sandbo x plan f or a limited time under regulator y super vision;

− (56) ‘ AI literacy’ means skills, knowledg e and understanding that allow provider s, deplo y ers and affect ed persons, taking

− into account their respective r ights and obliga tions in the context of this Regulation, to make an inf or med deplo yment

− of AI systems, as well as to g ain awareness about the oppor tunities and r isks of AI and possible har m it can cause;

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 49/144

− (57) ‘te sting in real-world conditions’ means the tem porar y t esting of an AI syste m f or its intended pur pose in real-wo rld

− conditions outside a laborato r y or other wise simulated envir onment, with a view to g ather ing reliable and robust data

− and to assessing and ver ifying the conf or mity of the AI system with the requirements of this Regulation and it does

− not qualify as placing the AI system on the marke t or putting it into ser vice within the meaning of this Regulation,

− pro vided that all the conditions laid do wn in Ar ticle 57 or 60 are fulf illed;

− (58) ‘subject’, f or the pur pose of real-world te sting, means a natural person who par ticipate s in te sting in real-wo rld

− conditions;

− (59) ‘inf or med consent’ means a subject’s freely given, specific, unambiguous and v oluntar y expression of his or her

− willingness to par ticipate in a par ticular testing in real-world conditions, af ter having been inf or med of all aspects of

− the te sting that are relevant to the subject ’s decision to par ticipate ;

− (60) ‘deep f ak e ’ means AI-g enerated or manipulate d imag e, audio or video cont ent that resembles existing persons, objects,

− places, entities or events and would f alsely appear to a person to be authentic or tr uthful;

− (61) ‘widespread infr ingement’ means an y act or omission contrar y to Union law prot ecting the inte rest of individuals,

− which :

− (a) has har med or is like ly to har m the collective interests of individuals residing in at least tw o Member States other

− than the Member State in which:

− (i) the act or omission or iginated or t ook place;

− (ii) the provider concer ned, or , where applicable, its author ised representative is located or established; or

− (iii) the deplo y er is established, when the infr ingement is committed b y the deplo y er;

− (b) has caused, causes or is likel y to cause har m to the collective intere sts of individuals and has common features,

− including the same unlawful practice or the same interest being infr inged, and is occur r ing concur rently ,

− committ ed by the same operat or , in at least three Member States;

− (62) ‘cr itical infrastr ucture’ means cr itical infrastr ucture as def ined in Ar ticle 2, point (4), of Directive (EU) 2022/2557;

− (63) ‘g eneral-pur pose AI model’ means an AI model, including where suc h an AI model is trained with a larg e amount of

− data using self-super vision at scale, that displa ys significant g enerality and is capable of compet ently perfo r ming

− a wide rang e of distinct tasks regardless of the wa y the model is placed on the market and that can be integrat ed into

− a var iety of do wnstream systems or applications, ex cept AI models that are used f or research, development or

− prot otyping activities bef ore the y are placed on the mark et ;

− (64) ‘high-imp act capabilities ’ means capabilities that matc h or ex ceed the capabilities recorded in the most advanced

− g eneral-pur pose AI models;

− (65) ‘syst emic r isk’ means a r isk that is specif ic to the high-im pact capabilities of g eneral-pur pose AI models, having

− a significant im pact on the Uni on market due t o their reac h, or due to actual or reasonably f oreseeable neg ative effe cts

− on public health, safety , public secur ity , fundamental r ights, or the society as a whole, that can be propagat ed at scale

− across the value ch ain;

− (66) ‘g eneral-pur pose AI syste m’ means an AI syste m which is based on a g eneral-pur pose AI model and which has the

− capability to ser ve a var iety of pur poses, both f or direct use as well as f or integrat ion in other AI systems;

− (67) ‘f loating-point operation’ means any mathematical operation or assignment in volving f loating-point numbers, which

− are a subset of the real numbers typically represente d on comput ers by an integ er of fixed precision scaled by an

− inte ger exponent of a fi xe d base;

− (68) ‘do wnstream pro vider ’ means a provid er of an AI system, including a g eneral-pur pose AI system, which integrat es an

− AI model, rega rdless of whether the AI model is provided by themselves and ver tically integrat ed or provid ed by

− another entity based on contractual relations.

− 50/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− Ar ticle 4

− AI literacy

− Provi ders and deplo y ers of AI syste ms shall take measures to ensure, to their best exte nt, a suffi cient level of AI lite racy of

− their staff and other persons dealing with the operation and use of AI systems on their behalf, taking into account their

− te chnical kno wledge, exper ience, education and training and the context the AI systems are t o be used in, and consider ing

− the persons or groups of persons on whom the AI syste ms are to be used.

− C HAPTER II

− PR OHIBITED AI PRA CTICES

− Ar ticle 5

− Prohibited AI practices

− 1. The f ollowi ng AI practices shall be prohibite d:

− (a) the placing on the market, the putting into ser vice or the use of an AI syste m that deplo ys subliminal tec hniques bey ond

− a person’s consciousness or pur posefully manipulative or decep tive t echniques, with the objective, or the eff ect of

− materially distor ting the behaviour of a person or a group of persons by appreciably imp air ing their ability to make an

− inf or med decision, thereby causing them to take a decision that the y would not hav e other wise take n in a manner that

− causes or is reasonably like ly to cause that person, another person or group of persons significant har m;

− (b) the placing on the marke t, the putting into ser vice or the use of an AI system that exploits any of the vulnerabilities of

− a natural person or a specif ic group of persons due to their age, disability or a specif ic social or economic situation, with

− the objective, or the effe ct, of materi ally distor ting the behavio ur of that person or a person belonging to that group in

− a manner that causes or is reasonably like ly to cause that person or another person significant har m;

− (c) the placing on the market, the putting into ser vice or the use of AI syste ms f or the evaluation or classif ication of natural

− persons or groups of persons o ver a cer tain per iod of time based on their social behavio ur or known, inferr ed or

− predicted personal or personality ch aracteristics , with the social score leading to either or both of the f ollo wing:

− (i) detr imental or unfa vourable treatment of cer tain natural persons or groups of persons in social contexts that are

− unrelat ed to the cont exts in which the data was or iginally generat ed or collect ed;

− (ii) detr imental or unfa v ourable treatment of cer tain natural persons or groups of persons that is unjustifie d or

− dispropor tionate to their social behavio ur or its gra vity ;

− (d) the placing on the marke t, the putting into ser vice f or this specif ic pur pose, or the use of an AI system f or making r isk

− assessments of natural persons in order to assess or predict the r isk of a natural person committing a cr iminal offe nce,

− based solely on the profiling of a natural person or on assessing their personality traits and ch aracter istics; this

− prohibition shall not apply to AI systems used to suppor t the human assessment of the in volvement of a person in

− a cr iminal activity , which is already based on objective and ver ifiable f acts directly linke d t o a cr iminal activity ;

− (e) the placing on the mark et, the putting into ser vice f or this specif ic pur pose, or the use of AI syste ms that create or

− expand f acial recognition databases through the untarget ed scraping of f acial images from the inter net or CCT V f ootage;

− (f) the placing on the marke t, the putting into ser vice f or this specific pur pose, or the use of AI syste ms to infer emotions

− of a natural person in the areas of wo rkplace and education institutions, ex cept where the use of the AI system is

− intende d to be put in place or into the mark et f or medical or saf ety reasons;

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 51/144

− (g) the placing on the marke t, the putting into ser vice f or this specific pur pose, or the use of biometr ic categor isation

− syste ms that catego r ise individually natural persons based on their biometr ic data to deduce or infer their race, political

− opinions, trade union membership, religious or philosophical beliefs, sex life or sexual or ientation; this prohibition does

− not cover any labelling or filter ing of law fully acquired biometr ic datasets, suc h as imag es, based on biometr ic data or

− categori zing of biometr ic data in the area of law enf orcement ;

− (h) the use of ‘real-time’ remote biometr ic identification syste ms in publicly accessible spaces f or the pur poses of la w

− enf orcement, unless and in so f ar as suc h use is str ictly necessar y f or one of the f ollo wing objectives:

− (i) the targe ted search f or specif ic victims of abduction, trafficking in human beings or sexual exploitation of human

− beings, as well as the search f or missing persons;

− (ii) the prevention of a specif ic, substantial and imminent threat to the life or physica l saf ety of natural persons or

− a genuine and present or g enuine and f oreseeable threat of a te r ror ist attack;

− (iii) the localisation or identification of a person suspect ed of having committed a cr iminal offence, f or the pur pose of

− conducting a cr iminal invest igation or prosecution or ex ecuting a cr iminal penalty f or offe nces referred to in

− Annex II and punishable in the Member Stat e concer ned by a custodial sente nce or a det ention order f or

− a maximum per iod of at least f our y ears.

− P oint (h) of the fir st subparagraph is without prejudice t o Ar ticle 9 of Regulation (EU) 2016/679 f or the processing of

− biometr ic data f or pur poses other than law enf orcement.

− 2. The use of ‘real-time’ remote biometr ic identifi cation syste ms in publicly accessible spaces f or the pur poses of la w

− enf orcement f or any of the objectives refer red to in paragraph 1, fir st subparagraph, point (h), shall be deplo y ed f or the

− pur poses set out in that point only to confir m the identity of the specif ically targe ted individual, and it shall take into

− account the f ollo wing elements:

− (a) the nature of the situation giving r ise to the possible use, in par ticular the ser iousness, probability and scale of the har m

− that would be caused if the system were not used;

− (b) the consequences of the use of the system f or the r ights and freedoms of all persons concer ned, in par ticular the

− ser iousness, probability and scale of those consequences.

− In addition, the use of ‘real-time’ remot e biometr ic identification systems in publicly accessible spaces f or the pur poses of

− law enf orcement f or any of the objectives referred to in paragraph 1, first subparagraph, point (h), of this Ar ticle shall

− comply with necessar y and propor tionate safegua rds and conditions in relation to the use in accordance with the national

− law author ising the use thereof, in par ticular as regard s the te mporal, g eographic and personal limitations. The use of the

− ‘real-time’ remot e biometr ic identifica tion syste m in publicly accessible spaces shall be author ised only if the la w

− enf orcement author ity has complet ed a fundamental r ights im pact assessment as provid ed f or in Ar ticle 27 and has

− register ed the system in the EU database according to Ar ticle 49. Ho wever , in duly justified cases of urgency , the use of such

− syste ms ma y be commenced without the registration in the EU database, provid ed that suc h registration is complet ed

− without undue dela y .

− 3. For the pur poses of paragraph 1, first subparagraph, point (h) and paragraph 2, each use f or the pur poses of la w

− enf orcement of a ‘real-time’ remot e biometr ic identifica tion syste m in publicly accessible spaces shall be subject to a pr ior

− author isation granted b y a judicial author ity or an independent administrative author ity whose decision is binding of the

− Member State in which the use is to take place, issued upon a reasoned request and in accordance with the detailed r ules of

− national law referred to in paragraph 5. How ever , in a duly justified situation of urg ency , the use of suc h system ma y be

− commenced without an author isation provided that suc h author isation is request ed without undue dela y , at the lat est

− within 24 hours. If such author isation is rejected, the use shall be st opped with immediate eff ect and all the data, as well as

− the results and outputs of that use shall be immediately discarded and deleted.

− The comp etent judicial author ity or an independent administrative author ity whose decision is binding shall grant the

− author isation only where it is satisfied, on the basis of objective evidence or clear indications presented to it, that the use of

− the ‘real-time’ remote biometr ic identifi cation system concer ned is necessar y f or , and propor tionate to , ac hieving one of the

− 52/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− objectives specified in paragraph 1, first subparagraph, point (h), as identified in the request and, in par ticular , remains

− limited to what is str ictly necessar y concer ning the per iod of time as well as the g eographic and personal scope. In deciding

− on the request, that author ity shall tak e into account the elements refe r red to in paragraph 2. No decision that produces an

− adverse leg al effe ct on a person ma y be tak en based solely on the output of the ‘real-time’ remote biometr ic identification

− syste m.

− 4. Without prejudice to paragraph 3, each use of a ‘real-time’ remot e biometr ic identifica tion system in publicly

− accessible spaces f or law enf orcement pur poses shall be notif ied to the relevant marke t sur veillance author ity and the

− national data prot ection author ity in accordance with the national r ules refer red t o in paragraph 5. The notif ication shall, as

− a minimum, contain the inf or mation specified under paragraph 6 and shall not include sensitive operational data.

− 5. A Member Stat e ma y decide t o provide f or the possibility to fully or par tially author ise the use of ‘real-time’ remote

− biometr ic identifica tion systems in publicly accessible spaces f or the pur poses of law enf orcement within the limits and

− under the conditions listed in paragraph 1, first subparagraph, point (h), and paragraphs 2 and 3. Member Stat es concer ned

− shall la y do wn in their national law the necessar y detailed r ules f or the request, issuance and ex ercise of, as well as

− super vision and repor ting relating to, the author isations refe r red to in paragraph 3. Those r ules shall also specify in respect

− of which of the objectives listed in paragraph 1, fir st subparagraph, point (h), including which of the cr iminal offe nces

− referred to in point (h)(iii) thereof, the compet ent author ities ma y be author ised to use those syste ms f or the pur poses of

− law enf orcement. Member States shall notify those r ules t o the Commission at the latest 30 da ys f ollo wing the adop tion

− thereof. Member States ma y introduce, in accordance with Uni on law , more restr ictive laws on the use of remote biometr ic

− identifica tion systems.

− 6. National market sur veillance author ities and the national data prot ection author ities of Member Stat es that have been

− notif ied of the use of ‘real-time’ remote biometr ic identifi cation syste ms in publicly accessible spaces f or law enf orcement

− pur poses pursuant to paragraph 4 shall submit to the Commission annual repor ts on suc h use. For that pur pose, the

− Commission shall pro vide Member States and national market sur veillance and data protection author ities with a tem plate ,

− including inf or mation on the number of the decisions taken by compet ent judicial author ities or an independent

− administrative author ity whose decision is binding upon requests f or author isations in accordance with paragraph 3 and

− their result.

− 7. The Commission shall publish annual repor ts on the use of real-time remot e biometr ic identification systems in

− publicly accessible spaces f or law enf orcement pur poses, based on aggregat ed data in Member States on the basis of the

− annual repor ts refer red to in paragraph 6. Those annual repor ts shall not include sensitive operational data of the related

− law enf orcement activities.

− 8. This Ar ticle shall not affe ct the prohibitions that apply where an AI practice infr ing es other Union law .

− CHAPTER III

− HIGH-RISK AI SYSTEMS

− SECTION 1

− Classification of AI sy stems as high-r isk

− Ar ticle 6

− Classif ication r ules f or high-r isk AI sys tems

− 1. Ir respective of whether an AI syste m is placed on the marke t or put into ser vice independently of the products

− referred to in points (a) and (b), that AI system shall be considered to be high-r isk where both of the f ollo wing conditions

− are fulfilled:

− (a) the AI syste m is inte nded to be used as a saf ety compo nent of a product, or the AI syste m is itself a product, cover ed by

− the Uni on har monisation legislation listed in Annex I;

− (b) the product whose safety compo nent pursuant to point (a) is the AI syste m, or the AI system itself as a product, is

− required to undergo a third-par ty conf or mity assessment, with a view to the placing on the marke t or the putting into

− ser vice of that product pursuant t o the Uni on har monisation legislation listed in Annex I.

− 2. In addition to the high-r isk AI syste ms refe r red to in paragraph 1, AI syste ms refe r red to in Annex III shall be

− considered to be high-r isk.

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 53/144

− 3. By derogat ion from paragraph 2, an AI system refe r red to in Annex III shall not be considered t o be high-r isk where it

− does not pose a significant r isk of har m to the health, safety or fundamental r ights of natural persons, including by not

− materially inf luencing the outcome of decision making.

− The first subparagraph shall apply where an y of the f ollo wing conditions is fulf illed:

− (a) the AI syste m is intende d to per f or m a nar row procedural task;

− (b) the AI syste m is intende d to impro ve the result of a previously comp let ed human activity ;

− (c) the AI system is intende d to det ect decision-making patt er ns or deviations from pr ior decision-making patterns and is

− not meant to replace or inf luence the previously comp leted human assessment, without proper human review ; or

− (d) the AI system is intended to perfor m a preparat or y task to an assessment relevant f or the pur poses of the use cases

− listed in Annex III.

− Notwithstanding the first subparagraph, an AI system refer red to in Annex III shall alwa ys be considered to be high-r isk

− where the AI system perfor ms profiling of natural persons.

− 4. A provid er who considers that an AI syste m refe r red to in Annex III is not high-r isk shall document its assessment

− bef ore that system is placed on the market or put into ser vice. Such provid er shall be subject to the registration obliga tion

− set out in Ar ticle 49(2). Upon request of national comp et ent author ities, the provider shall provide the documentation of

− the assessment.

− 5. The Commission shall, af te r consulting the European Ar tif icial Inte lligence Board (the ‘Board’), and no later than

− 2 Febr uar y 2026, provide guidelines specifying the practical im plementation of this Ar ticle in line with Ar ticle 96 to gether

− with a comp rehensive list of practical exam ples of use cases of AI systems that are high-r isk and not high-r isk.

− 6. The Commission is empo wered to adopt deleg ated acts in accordance with Ar ticle 97 in order to amend paragraph 3,

− second subparagraph, of this Ar ticle by adding new conditions to those laid do wn therein, or by modifying them, where

− there is concrete and reliable evidence of the existe nce of AI syste ms that f all under the scope of Annex III, but do not pose

− a significant r isk of har m to the health, saf ety or fundamental r ights of natural persons.

− 7. The Commission shall adop t delegat ed acts in accordance with Ar ticle 97 in order to amend paragraph 3, second

− subparagraph, of this Ar ticle by deleting any of the conditions laid down therein, where there is concrete and reliable

− evidence that this is necessar y t o maintain the level of protection of health, safety and fundamental r ights pro vided f or by

− this Regulation.

− 8. Any amendment to the conditions laid do wn in paragraph 3, second subparagraph, adopt ed in accordance with

− paragraphs 6 and 7 of this Ar ticle shall not decrease the overall level of protect ion of health, safety and fundamental r ights

− provid ed f or b y this Regulation and shall ensure consiste ncy with the deleg ated acts adop ted pursuant to Ar ticle 7(1), and

− take account of marke t and tec hnological developments.

− Ar ticle 7

− Amendments to Annex III

− 1. The Commission is em powered to adop t deleg ated acts in accordance with Ar ticle 97 t o amend Annex III b y adding

− or modifying use-cases of high-r isk AI systems where both of the f ollowing conditions are fulf illed:

− (a) the AI syste ms are intende d to be used in any of the areas list ed in Annex III;

− (b) the AI systems pose a r isk of har m to health and safety , or an adverse im pact on fundamental r ights, and that r isk is

− equivalent to, or great er than, the r isk of har m or of adverse imp act posed by the high-r isk AI syste ms already refe r red

− to in Annex III.

− 54/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− 2. When assessing the condition under paragraph 1, point (b), the Commission shall take into account the f ollowing

− cr ite r ia:

− (a) the inte nded pur pose of the AI system;

− (b) the exte nt to which an AI syste m has been used or is like ly t o be used;

− (c) the nature and amount of the data processed and used b y the AI system, in par ticular whether special categori es of

− personal data are processed;

− (d) the extent to which the AI system acts autonomously and the possibility f or a human to ove r r ide a decision or

− recommendations that ma y lead to pote ntial har m;

− (e) the exte nt to which the use of an AI syste m has already caused har m to health and safety , has had an adverse impact on

− fundamental r ights or has given r ise to significant concer ns in relation to the like lihood of suc h har m or adverse impact,

− as demonstrated, f or exam ple, by repor ts or documente d alleg ations submitted to national compet ent author ities or by

− other repor ts, as appropr iate;

− (f) the pote ntial exte nt of such har m or suc h adverse impact, in par ticular in ter ms of its intens ity and its ability t o affect

− multiple persons or to dispropor tionately affect a par ticular group of persons;

− (g) the extent to which persons who are potent ially har med or suffer an adverse impact are dependent on the outcome

− produced with an AI system, in par ticular because f or practical or lega l reasons it is not reasonably possible to opt-ou t

− from that outcome;

− (h) the extent to which there is an imbalance of power , or the persons who are potentially har med or suff er an adverse

− imp act are in a vulnerable position in relation to the deplo y er of an AI syste m, in par ticular due to status, author ity ,

− kno wledge, economic or social circumstances, or age;

− (i) the exte nt to which the outcome produced inv olving an AI system is easily cor r igible or reversible, taking into account

− the te chnical solutions available to cor rect or reverse it, whereby outcomes hav ing an adverse imp act on health, saf ety or

− fundamental r ights, shall not be considered to be easily cor r igible or reversible;

− (j) the magnitude and likelihood of benefit of the deplo yment of the AI system f or individuals, groups, or society at larg e,

− including possible im provements in product saf ety ;

− (k) the exte nt to which existing Uni on law provid es f or:

− (i) eff ective measures of redress in relation t o the r isks posed by an AI syste m, with the ex clusion of claims f or

− damage s;

− (ii) eff ective measures to prevent or substantially minimise those r isks.

− 3. The Commission is empo wered to adopt deleg ated acts in accordance with Ar ticle 97 t o amend the list in Annex III

− by removi ng high-r isk AI systems where both of the f ollowing conditions are fulf illed:

− (a) the high-r isk AI syste m concer ned no longer poses any signifi cant r isks to fundamental r ights, health or safety , taking

− into account the cr ite r ia listed in paragraph 2;

− (b) the deletion does not decrease the ove rall level of prot ection of health, saf ety and fundamental r ights under Union law .

− SECTION 2

− Requir ements f or high-r isk AI systems

− Ar ticle 8

− Compliance with the requirements

− 1. High-r isk AI systems shall comp ly with the requirements laid do wn in this Section, taking into account their intended

− pur pose as well as the g enerally ackno wledged stat e of the ar t on AI and AI-related tec hnologies. The r isk management

− syste m refe r red t o in Ar ticle 9 shall be taken into account when ensur ing compliance with those requirements.

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 55/144

− 2. Where a product contains an AI syste m, to which the requirements of this Regulation as well as requirements of the

− Uni on har monisation legislation listed in Section A of Annex I apply , provid ers shall be responsible f or ensur ing that their

− product is fully comp liant with all applicable requirements under applicable Uni on har monisation legislation. In ensur ing

− the comp liance of high-r isk AI systems refe r red t o in paragraph 1 with the requirements set out in this Section, and in order

− to ensure consistency , av oid duplication and minimise additional burdens, provid ers shall have a ch oice of integrat ing, as

− appropr iate, the necessar y te sting and repor ting processes, inf or mation and documentation they provide with rega rd to

− their product into documentation and procedures that already exist and are required under the Uni on har monisation

− legislation listed in Section A of Annex I.

− Ar ticle 9

− Risk management sys tem

− 1. A r isk managem ent system shall be established, imp lemented, documented and maintained in relation to high-r isk AI

− syste ms.

− 2. The r isk managem ent system shall be understood as a continuous iterative process planned and r un throughout the

− entire lif ecy cle of a high-r isk AI system, requir ing regular systematic review and updating. It shall comp r ise the f ollowing

− st eps:

− (a) the identification and analysis of the known and the reasonably f oreseeable r isks that the high-r isk AI system can pose

− to health, saf ety or fundamental r ights when the high-r isk AI syste m is used in accordance with its intende d pur pose;

− (b) the estimation and evaluation of the r isks that ma y emerg e when the high-r isk AI syste m is used in accordance with its

− intende d pur pose, and under conditions of reasonably f oreseeable misuse;

− (c) the evaluation of other r isks possibly ar ising, based on the analysis of data g athered from the post-mark et monitoring

− syste m refe r red t o in Ar ticle 72;

− (d) the adoption of appropr iate and target ed r isk management measures designed to address the r isks identified pursuant to

− point (a).

− 3. The r isks referred to in this Ar ticle shall concer n only those which ma y be reasonably mitiga te d or eliminated through

− the development or design of the high-r isk AI syste m, or the provision of adequat e te chnical inf or mation.

− 4. The r isk managem ent measures refe r red to in paragraph 2, point (d), shall give due consideration to the eff ects and

− possible inte raction resulting from the combined application of the requirements set out in this Section, with a view to

− minimising r isks more effectively while ac hieving an appropr iate balance in implementing the measures to fulfil those

− requirements.

− 5. The r isk managem ent measures refer red to in paragraph 2, point (d), shall be suc h that the relevant residual r isk

− associate d with each hazard, as well as the overall residual r isk of the high-r isk AI syste ms is judged to be accep table.

− In identifying the most appropr iate r isk managem ent measures, the f ollowing shall be ensured:

− (a) elimination or reduction of r isks identifie d and evaluate d pursuant to paragraph 2 in as far as tec hnically f easible

− through adequate design and development of the high-r isk AI syste m;

− (b) where appropr iate, imp lementation of adequate mitiga tion and control measures addressing r isks that cannot be

− eliminated ;

− (c) provision of inf or mation required pursuant to Ar ticle 13 and, where appropr iate, training to deplo y ers.

− With a view to eliminating or reducing r isks relate d to the use of the high-r isk AI system, due consideration shall be given

− to the tec hnical kno wledge, exper ience, education, the training to be expected b y the deplo y er , and the presumable cont ext

− in which the system is intended to be used.

− 56/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− 6. High-r isk AI systems shall be t ested f or the pur pose of identifying the most appropr iate and target ed r isk management

− measures. T esting shall ensure that high-r isk AI systems perfor m consist ently f or their intended pur pose and that they are in

− complia nce with the requirements set out in this Section.

− 7. T esting procedures ma y include t esting in real-world conditions in accordance with Ar ticle 60.

− 8. The testing of high-r isk AI syste ms shall be perfo r med, as appropr iate, at an y time throughout the development

− process, and, in any event, pr ior t o their being placed on the market or put into ser vice. T esting shall be car r ied out agains t

− pr ior def ined metr ics and probabilistic thresholds that are appropr iate to the intended pur pose of the high-r isk AI syste m.

− 9. When implementing the r isk management system as provid ed f or in paragraphs 1 t o 7, provider s shall give

− consideration to whether in view of its intende d pur pose the high-r isk AI syste m is like ly to hav e an adverse im pact on

− persons under the age of 18 and, as appropr iate, other vulnerable groups.

− 10. For provider s of high-r isk AI syste ms that are subject t o requirements regard ing inter nal r isk managem ent processes

− under other relevant provisions of Union law , the aspects provided in paragraphs 1 to 9 ma y be par t of, or combined with,

− the r isk managem ent procedures established pursuant to that law .

− Ar ticle 10

− Data and dat a go v er nance

− 1. High-r isk AI syste ms which mak e use of te ch niques in volving the training of AI models with data shall be developed

− on the basis of training, validation and testing data sets that meet the quality cr iter ia refe r red t o in paragraphs 2 to 5

− whenever suc h data sets are used.

− 2. T raining, validation and te sting data sets shall be subject to data gove r nance and management practices appropr iate

− f or the inte nded pur pose of the high-r isk AI syste m. Those practices shall concer n in par ticular:

− (a) the relevant design choices ;

− (b) data collection processes and the or igin of data, and in the case of personal data, the or iginal pur pose of the data

− collection;

− (c) relevant data-preparation processing operations, suc h as annotation, labelling, cleaning, updating, enr ic hment and

− aggregati on;

− (d) the f or mulation of assum ptions, in par ticular with respect t o the inf or mation that the data are supposed to measure and

− represent ;

− (e) an assessment of the ava ilability , quantity and suitability of the data sets that are needed;

− (f) examination in view of possible biases that are likel y to affect the health and saf ety of persons, hav e a negative im pact

− on fundamental r ights or lead t o discr imination prohibited under Uni on law , especially where data outputs inf luence

− input s f or future operations;

− (g) appropr iate measures t o detect, prevent and mitigat e possible biases identifie d according to point (f);

− (h) the identifica tion of relevant data gaps or shor tcomings that prevent compliance with this Regulation, and ho w those

− g aps and shor tcomings can be addressed.

− 3. T raining, validation and te sting data sets shall be relevant, suffici ently representative, and to the best exte nt possible,

− free of er rors and comp let e in view of the intended pur pose. They shall hav e the appropr iate statistical proper ties, including,

− where applicable, as regards the persons or groups of persons in relation t o whom the high-r isk AI system is intende d to be

− used. Those charact er istics of the data sets ma y be met at the level of individual data sets or at the level of a combination

− thereof.

− 4. Data sets shall take into account, to the extent required by the intended pur pose, the char acter istics or elements that

− are par ticular to the specific geographical, cont extual, behavi oural or functional setting within which the high-r isk AI

− syste m is intende d to be used.

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 57/144

− 5. T o the exte nt that it is str ictly necessar y f or the pur pose of ensur ing bias detection and cor rection in relation to the

− high-r isk AI systems in accordance with paragraph (2), points (f) and (g) of this Ar ticle, the provid ers of suc h systems ma y

− ex ceptionally process special categor ies of personal data, subject to appropr iate saf eguards f or the fundamental r ights and

− freedoms of natural persons. In addition to the provisions set out in Regulations (EU) 2016/679 and (EU) 2018/1725 and

− Directive (EU) 2016/680, all the f ollowing conditions must be met in order f or suc h processing to occur:

− (a) the bias det ection and cor rection cannot be effectively fulfilled b y processing other data, including synthetic or

− anonymised data;

− (b) the special cate gor ies of personal data are subject t o te chnical limitations on the re-use of the personal data, and

− state-of-t he-ar t secur ity and pr ivacy-preser ving measures, including pseudon ymisation;

− (c) the special cate gor ies of personal data are subject to measures to ensure that the personal data processed are secured,

− protect ed, subject t o suitable saf eguards, including str ict controls and documentation of the access, to av oid misuse and

− ensure that only author ised persons have access to those personal data with appropr iate conf identiality obliga tions;

− (d) the special categori es of personal data are not to be transmitte d, transferre d or other wise accessed by other par ties;

− (e) the special categor ies of personal data are delete d once the bias has been cor rect ed or the personal data has reac hed the

− end of its retention per iod, whichever comes fi rst ;

− (f) the records of processing activities pursuant to Regulations (EU) 2016/679 and (EU) 2018/1725 and Directive (EU)

− 2016/680 include the reasons why the processing of special categor ies of personal data was str ictly necessar y to det ect

− and cor rect biases, and why that objective could not be achi eved b y processing other data.

− 6. For the development of high-r isk AI syste ms not using t echniques inv olving the training of AI models, paragraphs 2

− to 5 apply only to the testing data sets.

− Ar ticle 11

− T echnical documentation

− 1. The t echnical documentation of a high-r isk AI syste m shall be drawn up bef ore that syste m is placed on the market or

− put into ser vice and shall be kep t up-to date.

− The te chnical documentation shall be drawn up in suc h a wa y as to demonstrate that the high-r isk AI system complie s with

− the requirements set out in this Section and t o provide national comp etent author ities and notified bodies with the

− necessar y inf or mation in a clear and compre hensive f or m t o assess the compliance of the AI syste m with those

− requirements. It shall contain, at a minimum, the elements set out in Annex IV . SMEs, including star t-ups, ma y provid e the

− elements of the tec hnical documentation specified in Annex IV in a simplified manner . T o that end, the Commission shall

− establish a simplified te chnical documentation f or m target ed at the needs of small and microent er pr ises. Where an SME,

− including a star t-up, opts to provide the inf or mation required in Annex IV in a sim plif ied manner , it shall use the f or m

− referred to in this paragraph. Notified bodies shall accep t the f or m f or the pur poses of the conf or mity assessment.

− 2. Where a high-r isk AI system related to a product covered b y the Uni on har monisation legislation listed in Section

− A of Annex I is placed on the marke t or put into ser vice, a sing le set of te chnical documentation shall be dra wn up

− containing all the inf or mation set out in paragraph 1, as well as the inf or mation required under those lega l acts.

− 3. The Commission is empo wered to adop t delegat ed acts in accordance with Ar ticle 97 in order t o amend Annex IV ,

− where necessar y , to ensure that, in light of tec hnical progress, the te chnical documentation provides all the inf or mation

− necessar y t o assess the compliance of the system with the requirements set out in this Section.

− 58/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− Ar ticle 12

− Record-ke eping

− 1. High-r isk AI syste ms shall te chnically allow f or the auto matic recording of events (logs) over the lifetime of the

− syste m.

− 2. In order to ensure a level of traceability of the functioning of a high-r isk AI system that is appropr iate to the intended

− pur pose of the syste m, logging capabilities shall enable the recording of events relevant f or:

− (a) identifying situations that ma y result in the high-r isk AI system presenting a r isk within the meaning of Ar ticle 79(1) or

− in a substantial modification;

− (b) f acilitating the post-market monitoring refer red to in Ar ticle 72; and

− (c) monitoring the operation of high-r isk AI systems refe r red to in Ar ticle 26(5).

− 3. For high-r isk AI systems refe r red to in point 1 (a), of Annex III, the logging capabilities shall provid e, at a minimum:

− (a) recording of the per iod of each use of the syste m (star t date and time and end date and time of each use);

− (b) the refe rence database aga inst which in put data has been ch eck ed by the syste m;

− (c) the in put data f or which the search has led to a matc h;

− (d) the identification of the natural persons in v olved in the ver ification of the results, as refer red to in Ar ticle 14(5).

− Ar ticle 13

− T ransparency and pro vision of inf or mation t o deplo y ers

− 1. High-r isk AI systems shall be designed and developed in suc h a wa y as to ensure that their operation is suff iciently

− transparent to enable deplo y ers to inte r pret a syste m’s output and use it appropr iat ely . An appropr iate type and degree of

− transparency shall be ensured with a view to ac hieving comp liance with the relevant obliga tions of the provid er and

− deplo yer set out in Section 3.

− 2. High-r isk AI systems shall be accompanied by instr uctions f or use in an appropr iate digital f or mat or other wise that

− include concise, complet e, cor rect and clear inf or mation that is relevant, accessible and comp rehensible to deplo y ers.

− 3. The instr uctions f or use shall contain at least the f ollo wing inf or mation:

− (a) the identity and the contact details of the pro vider and, where applicable, of its author ised representative;

− (b) the ch aracter istics, capabilities and limitations of perf or mance of the high-r isk AI syste m, including:

− (i) its intended pur pose;

− (ii) the level of accuracy , including its metr ics, robustness and cybersecur ity refe r red to in Ar ticle 15 ag ainst which the

− high-r isk AI syste m has been test ed and validat ed and which can be expected, and any known and f oreseeable

− circumstances that ma y hav e an imp act on that expected level of accuracy , robustness and cybersecur ity ;

− (iii) an y kno wn or f oreseeable circumstance, relate d to the use of the high-r isk AI syste m in accordance with its

− inte nded pur pose or under conditions of reasonably f oreseeable misuse, which ma y lead t o r isks to the health and

− saf ety or fundamental r ights refer red to in Ar ticle 9(2);

− (iv) where applicable, the t echnical capabilities and ch aracter istics of the high-r isk AI system to provid e inf or mation

− that is relevant t o explain its output ;

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 59/144

− (v) when appropr iate, its perform ance rega rding specific persons or groups of persons on which the system is

− inte nded to be used;

− (vi) when appropr iate, specif ications f or the in put data, or any other relevant inf or mation in ter ms of the training,

− validation and testing data sets used, taking into account the intende d pur pose of the high-r isk AI system;

− (vii) where applicable, inf or mation to enable deplo y ers to inter pret the output of the high-r isk AI system and use it

− appropr iat ely ;

− (c) the chang es to the high-r isk AI syste m and its perf or mance which hav e been pre-dete r mined by the pro vider at the

− moment of the initial conf or mity assessment, if any ;

− (d) the human ove rsight measures referred to in Ar ticle 14, including the te ch nical measures put in place to f acilitate the

− inter pretation of the outputs of the high-r isk AI syste ms by the deplo y ers;

− (e) the computa tional and hardware resources needed, the expect ed lif etime of the high-r isk AI system and any necessar y

− mainte nance and care measures, including their frequency , to ensure the proper functioning of that AI system, including

− as regards sof tware update s;

− (f) where relevant, a descr iption of the mechanisms included within the high-r isk AI system that allows deplo y ers to

− properly collect, st ore and interpret the logs in accordance with Ar ticle 12.

− Ar ticle 14

− Human o v ersight

− 1. High-r isk AI systems shall be designed and developed in suc h a wa y , including with appropr iate human-machi ne

− interfac e too ls, that they can be effe ctively oversee n b y natural persons dur ing the per iod in which they are in use.

− 2. Human ove rsight shall aim to prevent or minimise the r isks t o health, saf ety or fundamental r ights that ma y emerge

− when a high-r isk AI syste m is used in accordance with its inte nded pur pose or under conditions of reasonably f oreseeable

− misuse, in par ticular where suc h r isks persist despite the application of other requirements set out in this Section.

− 3. The overs ight measures shall be commensurat e with the r isks, level of auto nomy and context of use of the high-r isk

− AI system, and shall be ensured through either one or both of the f ollowing types of measures:

− (a) measures identified and built, when te chnically feasible, into the high-r isk AI system by the provid er bef ore it is placed

− on the mark et or put into ser vice;

− (b) measures identifie d by the provider bef ore placing the high-r isk AI syste m on the marke t or putting it into ser vice and

− that are appropr iate t o be implement ed b y the deplo y er .

− 4. For the pur pose of imp lementing paragraphs 1, 2 and 3, the high-r isk AI system shall be provided to the deplo yer in

− suc h a wa y that natural persons to whom human oversight is assigned are enabled, as appropr iat e and propor tionate :

− (a) to properly understand the relevant capacities and limitations of the high-r isk AI system and be able to duly monitor its

− operation, including in view of detecting and addressing anomalies, dysfunctions and unexpect ed perfo r mance;

− (b) to remain aw are of the possible tendency of automat ically relying or over -relying on the output produced by a high-r isk

− AI syste m (automation bias), in par ticular f or high-r isk AI systems used to provide inf or mation or recommendations f or

− decisions to be take n b y natural persons;

− (c) to cor rectly inte r pret the high-r isk AI system’s output, taking into account, f or exam ple, the interpretat ion tools and

− methods available;

− 60/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− (d) to decide, in any par ticular situation, not to use the high-r isk AI syste m or to other wise disreg ard, overr ide or reverse

− the output of the high-r isk AI syste m;

− (e) to inter vene in the operation of the high-r isk AI syste m or inter r upt the syste m through a ‘st op’ button or a similar

− procedure that allows the syste m to come to a halt in a saf e state.

− 5. For high-r isk AI syste ms refer red to in point 1(a) of Annex III, the measures referred t o in paragraph 3 of this Ar ticle

− shall be suc h as to ensure that, in addition, no action or decision is take n b y the deplo y er on the basis of the identification

− resulting from the system unless that identification has been separate ly ver ified and confi r med b y at least tw o natural

− persons with the necessar y compet ence, training and author ity .

− The requirement f or a separate ver ificati on by at least tw o natural persons shall not apply to high-r isk AI syste ms used f or

− the pur poses of law enf orcement, migration, border control or asylum, where Union or national law considers the

− application of this requirement to be dispropor tionate .

− Ar ticle 15

− A ccuracy , robustness and cybersecur ity

− 1. High-r isk AI syste ms shall be designed and developed in suc h a wa y that the y ac hieve an appropr iate level of accuracy ,

− robustness, and cybersecur ity , and that the y perf or m consiste ntly in those respects throughout their lifecy cle.

− 2. T o address the t echnical aspects of ho w to measure the appropr iate levels of accuracy and robustness set out in

− paragraph 1 and any other relevant perform ance metr ics, the Commission shall, in cooperation with relevant stakeholders

− and organisations such as metrology and bench marking author ities, encourage , as appropr iate, the development of

− bench mark s and measurement methodologies.

− 3. The levels of accuracy and the relevant accuracy metr ics of high-r isk AI syste ms shall be declared in the accom panying

− instr uctions of use.

− 4. High-r isk AI syste ms shall be as resilient as possible regard ing er rors, f aults or inconsiste ncies that ma y occur within

− the syste m or the environment in which the syste m operat es, in par ticular due to their interaction with natural persons or

− other syste ms. T echnical and org anisational measures shall be take n in this regard .

− The robustness of high-r isk AI syste ms ma y be ac hieved through te chnical redundancy solutions, which ma y include

− backu p or f ail-safe plans.

− High-r isk AI syste ms that continue to lear n af ter being placed on the market or put into ser vice shall be developed in such

− a wa y as to eliminate or reduce as far as possible the r isk of possibly biased outputs inf luencing in put f or future operations

− (f eedback loops), and as to ensure that any suc h f eedback loops are duly addressed with appropr iate mitigati on measures.

− 5. High-r isk AI systems shall be resilient against atte mp ts by unauthor ised third par ties to alter their use, outputs or

− perfo r mance b y exploiting syste m vulnerabilities.

− The tec hnical solutions aiming to ensure the cybersecur ity of high-r isk AI syste ms shall be appropr iate t o the relevant

− circumstances and the r isk s.

− The tec hnical solutions to address AI specific vulnerabilities shall include, where appropr iate, measures to prevent, detect,

− respond to, resolve and control f or attac ks tr ying to manipulat e the training data set (data poisoning), or pre-trained

− compo nents used in training (model poisoning), in puts designed to cause the AI model t o make a mistake (adversar ial

− exam ples or model evasion), confi dentiality attac k s or model f la ws.

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 61/144

− SECTION 3

− Oblig ations of pr o viders and deploye rs of high-r isk AI syste ms and other par ties

− Ar ticle 16

− Obligations of pro viders of high-r isk AI systems

− Provi ders of high-r isk AI syste ms shall:

− (a) ensure that their high-r isk AI systems are compliant with the requirements set out in Section 2;

− (b) indicate on the high-r isk AI syste m or , where that is not possible, on its packaging or its accom panying documentation,

− as applicable, their name, regist ered trade name or register ed trade mark , the address at which the y can be contacted;

− (c) hav e a quality management syste m in place which comp lies with Ar ticle 17;

− (d) keep the documentation refer red to in Ar ticle 18;

− (e) when under their control, keep the logs auto matically g enerated by their high-r isk AI systems as refe r red t o in

− Ar ticle 19;

− (f) ensure that the high-r isk AI system undergoes the relevant conf or mity assessment procedure as refer red to in Ar ticle 43,

− pr ior to its being placed on the mark et or put into ser vice;

− (g) dra w up an EU declaration of conf or mity in accordance with Ar ticle 47;

− (h) affix the CE marking to the high-r isk AI system or , where that is not possible, on its packag ing or its accom pan ying

− documentation, to indicate conf or mity with this Regulation, in accordance with Ar ticle 48;

− (i) comply with the registration oblig ations refer red to in Ar ticle 49(1);

− (j) take the necessar y cor rective actions and pro vide inf or mation as required in Ar ticle 20;

− (k) upon a reasoned request of a national comp etent author ity , demonstrate the conf or mity of the high-r isk AI system with

− the requirements set out in Section 2;

− (l) ensure that the high-r isk AI system com plies with accessibility requirements in accordance with Directives (EU)

− 2016/2102 and (EU) 2019/882.

− Ar ticle 17

− Quality management sys tem

− 1. Providers of high-r isk AI syste ms shall put a quality manag ement system in place that ensures comp liance with this

− Regulation. That system shall be documente d in a syste matic and orderly manner in the f or m of wr itten policies, procedures

− and instr uctions, and shall include at least the f ollowing aspects:

− (a) a strate gy f or regulator y comp liance, including comp liance with conf or mity assessment procedures and procedures f or

− the manag ement of modifications to the high-r isk AI system;

− (b) te chniques, procedures and syste matic actions to be used f or the design, design control and design ver ification of the

− high-r isk AI syste m;

− (c) te chniques, procedures and systematic actions to be used f or the development, quality control and quality assurance of

− the high-r isk AI system;

− (d) examination, te st and validation procedures to be car r ied out bef ore, dur ing and af te r the development of the high-r isk

− AI system, and the frequency with which they hav e to be car r ied out ;

− 62/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− (e) te chnical specifications, including standards, to be applied and, where the relevant har monised standards are not

− applied in full or do not cover all of the relevant requirements set out in Section 2, the means to be used t o ensure that

− the high-r isk AI system complies with those requirements;

− (f) syste ms and procedures f or data managem ent, including data acquisition, data collection, data analysis, data labelling,

− data st orage, data filt ration, data mining, data aggregation, data ret ention and any other operation regard ing the data

− that is perfo r med bef ore and f or the pur pose of the placing on the mark et or the putting into ser vice of high-r isk AI

− syste ms;

− (g) the r isk managem ent syste m refe r red to in Ar ticle 9;

− (h) the setting-up, imp lementation and maintenance of a post-mark et monitoring syste m, in accordance with Ar ticle 72;

− (i) procedures related to the repor ting of a ser ious incident in accordance with Ar ticle 73;

− (j) the handling of communication with national compet ent author ities, other relevant author ities, including those

− provid ing or suppor ting the access to data, notif ied bodies, other operators, custome rs or other intere sted par ties;

− (k) syste ms and procedures f or record-keepi ng of all relevant documentation and inf or mation;

− (l) resource manag ement, including secur ity-of-supply related measures;

− (m) an accountability framework setting out the responsibilities of the management and other staff with regard to all the

− aspects listed in this paragraph.

− 2. The imp lementation of the aspects refer red to in paragraph 1 shall be propor tionate to the size of the provider ’s

− org anisation. Providers shall, in any event, respect the degree of r igour and the level of protection required to ensure the

− complia nce of their high-r isk AI systems with this Regulation.

− 3. Providers of high-r isk AI systems that are subject to obliga tions regard ing quality management systems or an

− equivalent function under relevant sectoral Uni on law ma y include the aspects listed in paragraph 1 as par t of the quality

− managem ent syste ms pursuant to that law .

− 4. For provid ers that are fi nancial institutions subject to requirements regard ing their inter nal gove r nance, ar rangements

− or processes under Uni on financial ser vices law , the obliga tion to put in place a quality management syste m, with the

− ex ception of paragraph 1, points (g), (h) and (i) of this Ar ticle, shall be deemed to be fulfilled by comp lying with the r ules on

− inter nal gover nance ar rang ements or processes pursuant to the relevant Uni on financ ial ser vices law . T o that end, any

− har monised standards refe r red to in Ar ticle 40 shall be take n into account.

− Ar ticle 18

− Documentation keep ing

− 1. The provider shall, f or a per iod ending 10 y ears af te r the high-r isk AI syste m has been placed on the marke t or put

− into ser vice, keep at the disposal of the national compet ent author ities:

− (a) the t echnical documentation refer red to in Ar ticle 11;

− (b) the documentation concer ning the quality management syste m refe r red t o in Ar ticle 17;

− (c) the documentation concer ning the ch anges approved by notif ied bodies, where applicable;

− (d) the decisions and other documents issued by the notified bodies, where applicable;

− (e) the EU declaration of conf or mity referred to in Ar ticle 47.

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 63/144

− 2. Each Member State shall det er mine conditions under which the documentation referred to in paragraph 1 remains at

− the disposal of the national compet ent author ities f or the per iod indicated in that paragraph f or the cases when a provid er

− or its author ised representative established on its terr itory goes bankr up t or ceases its activity pr ior to the end of that

− per iod.

− 3. Providers that are financial institutions subject to requirements regard ing their inter nal gover nance, ar rangements or

− processes under Union financ ial ser vices law shall maintain the te chnical documentation as par t of the documentation kep t

− under the relevant Uni on financial ser vices law .

− Ar ticle 19

− Aut omatically generated logs

− 1. Providers of high-r isk AI syste ms shall keep the logs refer red to in Ar ticle 12(1), automat ically g enerated by their

− high-r isk AI syste ms, t o the extent suc h logs are under their control. Without prejudice t o applicable Uni on or national law ,

− the logs shall be k ept f or a per iod appropr iate to the inte nded pur pose of the high-r isk AI syste m, of at least six months,

− unless provided other wise in the applicable Uni on or national la w , in par ticular in Union law on the protect ion of personal

− data.

− 2. Providers that are financial institutions subject to requirements regard ing their inter nal gover nance, ar rangements or

− processes under Union fi nancial ser vices law shall maintain the logs automa tically g enerated by their high-r isk AI syste ms

− as par t of the documentation kep t under the relevant financial ser vices law .

− Ar ticle 20

− Cor rectiv e actions and duty of infor mation

− 1. Providers of high-r isk AI syste ms which consider or hav e reason to consider that a high-r isk AI system that they have

− placed on the market or put into ser vice is not in conf or mity with this Regulation shall immediate ly take the necessar y

− cor rective actions to br ing that system into conf or mity , to withdraw it, to disable it, or to recall it, as appropr iate. They shall

− inf or m the distr ibutors of the high-r isk AI syste m concer ned and, where applicable, the deplo yers, the author ised

− representative and imp or te rs according ly .

− 2. Where the high-r isk AI system presents a r isk within the meaning of Ar ticle 79(1) and the provider becomes aware of

− that r isk, it shall immediate ly investig ate the causes, in collaboration with the repor ting deplo y er , where applicable, and

− inf or m the marke t sur veillance author ities compet ent f or the high-r isk AI system concer ned and, where applicable, the

− notif ied body that issued a cer tificate f or that high-r isk AI system in accordance with Ar ticle 44, in par ticular , of the nature

− of the non-comp liance and of any relevant cor rective action taken.

− Ar ticle 21

− Cooperation with competent author ities

− 1. Providers of high-r isk AI systems shall, upon a reasoned request by a comp et ent author ity , pro vide that author ity all

− the inf or mation and documentation necessar y to demonstrate the conf or mity of the high-r isk AI syste m with the

− requirements set out in Section 2, in a languag e which can be easily understood by the author ity in one of the off icial

− languag es of the institutions of the Union as indicate d by the Member State concer ned.

− 2. Upon a reasoned request by a compet ent author ity , providers shall also give the requesting comp et ent author ity , as

− applicable, access to the automat ically g enerated logs of the high-r isk AI syste m refer red to in Ar ticle 12(1), to the exte nt

− suc h logs are under their control.

− 3. Any inf or mation obtained by a comp etent author ity pursuant to this Ar ticle shall be treate d in accordance with the

− confi dentiality oblig ations set out in Ar ticle 78.

− 64/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− Ar ticle 22

− Author ised representativ es of pro viders of high-r isk AI systems

− 1. Pr ior to making their high-r isk AI syste ms ava ilable on the Union mark et, provider s established in third countr ies

− shall, by wr itt en mandate, appoint an author ised representative which is established in the Uni on.

− 2. The provid er shall enable its author ised representative to perf or m the tasks specified in the mandate received from the

− provid er .

− 3. The author ised representative shall perfo r m the tasks specified in the mandate received from the provider . It shall

− provid e a cop y of the mandate to the marke t sur veillance author ities upon request, in one of the off icial languages of the

− institutions of the Union, as indicated by the compet ent author ity . For the pur poses of this Regulation, the mandate shall

− emp ower the author ised representative to car r y out the f ollowing tasks :

− (a) ver ify that the EU declaration of conf or mity refer red to in Ar ticle 47 and the te chnical documentation referred to in

− Ar ticle 11 hav e been drawn up and that an appropr iat e conf or mity assessment procedure has been car r ied out by the

− provid er ;

− (b) keep at the disposal of the comp etent author ities and national author ities or bodies refe r red to in Ar ticle 74(10), f or

− a per iod of 10 y ears af te r the high-r isk AI syste m has been placed on the marke t or put into ser vice, the contact details

− of the provid er that appoint ed the author ised representative, a cop y of the EU declaration of conf or mity refer red to in

− Ar ticle 47, the te chnical documentation and, if applicable, the cer tificate issued by the notif ied body ;

− (c) provid e a compet ent author ity , upon a reasoned request, with all the inf or mation and documentation, including that

− referred to in point (b) of this subparagraph, necessar y to demonstrate the conf or mity of a high-r isk AI system with the

− requirements set out in Section 2, including access to the logs, as refe r red to in Ar ticle 12(1), auto matically g enerated by

− the high-r isk AI system, to the extent suc h logs are under the control of the provid er ;

− (d) cooperate with compet ent author ities, upon a reasoned request, in any action the latter take in relation to the high-r isk

− AI system, in par ticular to reduce and mitiga te the r isks posed by the high-r isk AI system;

− (e) where applicable, comply with the registration obliga tions refer red to in Ar ticle 49(1), or , if the registration is car r ied

− out by the provid er itself, ensure that the inf or mation refer red to in point 3 of Section A of Annex VIII is cor rect.

− The mandate shall empo wer the author ised representative to be addressed, in addition t o or instead of the provider , by the

− compet ent author ities, on all issues related to ensur ing comp liance with this Regulation.

− 4. The author ised representative shall te r minate the mandat e if it considers or has reason to consider the provider to be

− acting contrar y to its obliga tions pursuant to this Regulation. In suc h a case, it shall immediately inf or m the relevant marke t

− sur veillance author ity , as well as, where applicable, the relevant notified body , about the te r mination of the mandat e and the

− reasons theref or .

− Ar ticle 23

− Obligations of impor ters

− 1. Bef ore placing a high-r isk AI syste m on the marke t, imp or t ers shall ensure that the syste m is in conf or mity with this

− Regulation by ver ifying that:

− (a) the relevant conf or mity assessment procedure refer red to in Ar ticle 43 has been car r ied out by the provider of the

− high-r isk AI syste m;

− (b) the provid er has dra wn up the tec hnical documentation in accordance with Ar ticle 11 and Annex IV ;

− (c) the system bears the required CE marking and is accompanied b y the EU declaration of conf or mity refer red to in

− Ar ticle 47 and instr uctions f or use;

− (d) the provid er has appoint ed an author ised representative in accordance with Ar ticle 22(1).

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 65/144

− 2. Where an impor te r has sufficient reason to consider that a high-r isk AI system is not in conf or mity with this

− Regulation, or is f alsifi ed, or accompanie d by f alsifi ed documentation, it shall not place the syste m on the market until it has

− been brought into conf or mity . Where the high-r isk AI syste m presents a r isk within the meaning of Ar ticle 79(1), the

− imp or te r shall inf or m the provid er of the syste m, the author ised representative and the market sur veillance author ities to

− that eff ect.

− 3. Impor te rs shall indicate their name, regist ered trade name or register ed trade mark, and the address at which they can

− be contacted on the high-r isk AI syste m and on its packaging or its accom pan ying documentation, where applicable.

− 4. Impor te rs shall ensure that, while a high-r isk AI syste m is under their responsibility , st orage or transpor t conditions,

− where applicable, do not jeopardise its compliance with the requirements set out in Section 2.

− 5. Impor te rs shall keep, f or a per iod of 10 y ears af ter the high-r isk AI system has been placed on the mark et or put into

− ser vice, a cop y of the cer tif icat e issued by the notif ied body , where applicable, of the instr uctions f or use, and of the EU

− declaration of conf or mity refe r red t o in Ar ticle 47.

− 6. Impor te rs shall provid e the relevant comp etent author ities, upon a reasoned request, with all the necessar y

− inf or mation and documentation, including that refer red to in paragraph 5, to demonstrat e the conf or mity of a high-r isk AI

− syste m with the requirements set out in Section 2 in a languag e which can be easily understoo d by them. For this pur pose,

− they shall also ensure that the tec hnical documentation can be made a vailable to those author ities.

− 7. Impor te rs shall cooperate with the relevant compet ent author ities in any action those author ities take in relation to

− a high-r isk AI syste m placed on the marke t by the im por ters, in par ticular t o reduce and mitig ate the r isks posed by it.

− Ar ticle 24

− Obligations of distr ibutors

− 1. Bef ore making a high-r isk AI system a vailable on the marke t, distr ibutors shall ver ify that it bears the required CE

− marking, that it is accom panied by a cop y of the EU declaration of conf or mity refe r red to in Ar ticle 47 and instr uctions f or

− use, and that the provid er and the imp or te r of that syste m, as applicable, hav e complied with their respective obliga tions as

− laid down in Ar ticle 16, points (b) and (c) and Ar ticle 23(3).

− 2. Where a distr ibutor considers or has reason to consider , on the basis of the inf or mation in its possession, that

− a high-r isk AI system is not in conf or mity with the requirements set out in Section 2, it shall not mak e the high-r isk AI

− syste m available on the market until the syste m has been brought into conf or mity with those requirements. Fur ther more,

− where the high-r isk AI system presents a r isk within the meaning of Ar ticle 79(1), the distr ibutor shall inf or m the provid er

− or the imp or te r of the syste m, as applicable, to that effe ct.

− 3. Distr ibut ors shall ensure that, while a high-r isk AI syste m is under their responsibility , st orage or transpor t

− conditions, where applicable, do not jeopardise the compliance of the syste m with the requirements set out in Section 2.

− 4. A distr ibut or that considers or has reason to consider , on the basis of the inf or mation in its possession, a high-r isk AI

− syste m which it has made a vailable on the marke t not t o be in conf or mity with the requirements set out in Section 2, shall

− take the cor rective actions necessar y to br ing that system into conf or mity with those requirements, to withdra w it or recall

− it, or shall ensure that the provid er , the impor te r or any relevant operat or , as appropr iat e, take s those cor rective actions.

− Where the high-r isk AI system presents a r isk within the meaning of Ar ticle 79(1), the distr ibut or shall immediate ly inf or m

− the provid er or impor te r of the syste m and the author ities compet ent f or the high-r isk AI system concer ned, giving details,

− in par ticular , of the non-com pliance and of any cor rective actions taken.

− 5. Upon a reasoned request from a relevant comp etent author ity , distr ibut ors of a high-r isk AI syste m shall provide that

− author ity with all the inf or mation and documentation regarding their actions pursuant to paragraphs 1 to 4 necessar y to

− demonstrate the conf or mity of that system with the requirements set out in Section 2.

− 6. Distr ibut ors shall cooperate with the relevant compet ent author ities in any action those author ities take in relation to

− a high-r isk AI system made a vailable on the market by the distr ibutors, in par ticular to reduce or mitigat e the r isk posed by

− it.

− 66/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− Ar ticle 25

− Responsibilities along the AI value chain

− 1. Any distr ibut or , imp or te r , deplo y er or other third-par ty shall be considered t o be a provid er of a high-r isk AI system

− f or the pur poses of this Regulation and shall be subject to the obliga tions of the provider under Ar ticle 16, in any of the

− f ollo wing circumstances:

− (a) they put their name or trademark on a high-r isk AI system already placed on the market or put into ser vice, without

− prejudice to contractual ar rangements stipulating that the obliga tions are other wise allocat ed;

− (b) they mak e a substantial modification to a high-r isk AI system that has already been placed on the market or has already

− been put into ser vice in suc h a wa y that it remains a high-r isk AI system pursuant to Ar ticle 6;

− (c) they modify the intended pur pose of an AI system, including a g eneral-pur pose AI system, which has not been classif ied

− as high-r isk and has already been placed on the marke t or put into ser vice in such a wa y that the AI syste m concer ned

− becomes a high-r isk AI system in accordance with Ar ticle 6.

− 2. Where the circumstances refe r red to in paragraph 1 occur , the provid er that initially placed the AI system on the

− marke t or put it into ser vice shall no longe r be considered to be a provider of that specif ic AI syste m f or the pur poses of

− this Regulation. That initial provider shall closely cooperate with new provid ers and shall make ava ilable the necessar y

− inf or mation and provide the reasonably expected t echnical access and other assistance that are required f or the fulf ilment of

− the oblig ations set out in this Regulation, in par ticular rega rding the compliance with the conf or mity assessment of

− high-r isk AI syste ms. This paragraph shall not apply in cases where the initial provider has clearly specif ied that its AI

− syste m is not to be ch anged into a high-r isk AI system and theref ore does not f all under the obligation to hand over the

− documentation.

− 3. In the case of high-r isk AI syste ms that are saf ety components of products co vered b y the Union har monisation

− legislation listed in Section A of Annex I, the product manufa cturer shall be considered to be the provider of the high-r isk

− AI system, and shall be subject to the obliga tions under Ar ticle 16 under either of the f ollo wing circumstances:

− (a) the high-r isk AI syste m is placed on the marke t to gether with the product under the name or trademark of the product

− manufa cturer;

− (b) the high-r isk AI system is put into ser vice under the name or trademark of the product manufacturer af te r the product

− has been placed on the marke t.

− 4. The provider of a high-r isk AI system and the third par ty that supplies an AI syste m, tools, ser vices, components, or

− processes that are used or integrat ed in a high-r isk AI system shall, by wr itten agreement, specify the necessar y inf or mation,

− capabilities, t echnical access and other assistance based on the g enerally ac knowledg ed stat e of the ar t, in order to enable

− the provider of the high-r isk AI syste m t o fully comply with the obliga tions set out in this Regulation. This paragraph shall

− not apply to third par ties making accessible to the public too ls, ser vices, processes, or comp onents, other than

− g eneral-pur pose AI models, under a free and open-source licence.

− The AI Office ma y develop and recommend v oluntar y model te r ms f or contracts between provid ers of high-r isk AI syste ms

− and third par ties that supply tools, ser vices, comp onents or processes that are used f or or inte grated into high-r isk AI

− syste ms. When developing those v oluntar y model ter ms, the AI Office shall take into account possible contractual

− requirements applicable in specif ic sect ors or business cases. The v oluntar y model ter ms shall be published and be ava ilable

− free of charg e in an easily usable electronic f or mat.

− 5. Paragraphs 2 and 3 are without prejudice to the need to obser ve and prot ect intellect ual proper ty r ights, confi dential

− business inf or mation and trade secrets in accordance with Uni on and national law .

− Ar ticle 26

− Obligations of deplo y ers of high-r isk AI systems

− 1. Deplo y ers of high-r isk AI systems shall tak e appropr iate te chnical and org anisational measures to ensure they use

− suc h syste ms in accordance with the instr uctions f or use accompan ying the systems, pursuant t o paragraphs 3 and 6.

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 67/144

− 2. Deplo y ers shall assign human ove rsight to natural persons who hav e the necessar y compet ence, training and

− author ity , as well as the necessar y suppor t.

− 3. The obligations set out in paragraphs 1 and 2, are without prejudice t o other deplo y er obligations under Uni on or

− national law and t o the deplo yer ’s freedom t o organise its own resources and activities f or the pur pose of imp lementing the

− human overs ight measures indicated by the provid er .

− 4. Without prejudice to paragraphs 1 and 2, to the extent the deplo y er exercises control over the input data, that

− deplo yer shall ensure that in put data is relevant and suffi ciently representative in view of the intende d pur pose of the

− high-r isk AI syste m.

− 5. Deplo y ers shall monitor the operation of the high-r isk AI syste m on the basis of the instr uctions f or use and, where

− relevant, inf or m provid ers in accordance with Ar ticle 72. Where deplo yers have reason to consider that the use of the

− high-r isk AI system in accordance with the instr uctions ma y result in that AI syste m presenting a r isk within the meaning of

− Ar ticle 79(1), they shall, without undue dela y , inf or m the provider or distr ibut or and the relevant market sur veillance

− author ity , and shall suspend the use of that system. Where deplo y ers have identifie d a ser ious incident, they shall also

− immediately inf or m first the provid er , and then the impor te r or distr ibut or and the relevant market sur veillance author ities

− of that incident. If the deplo y er is not able to reac h the provider , Ar ticle 73 shall apply mutatis mutandis. This obliga tion

− shall not cover sensitive operational data of deplo yers of AI systems which are law enf orcement author ities.

− For deplo y ers that are financ ial institutions subject to requirements regard ing their inter nal gover nance, ar rangements or

− processes under Union fi nancial ser vices law , the monitorin g obligation set out in the fir st subparagraph shall be deemed to

− be fulf illed by complying with the r ules on inter nal govern ance ar rangements, processes and mechanisms pursuant t o the

− relevant financial ser vice la w .

− 6. Deplo y ers of high-r isk AI syste ms shall k eep the logs auto matically g enerated b y that high-r isk AI system t o the exte nt

− suc h logs are under their control, f or a per iod appropr iate to the intended pur pose of the high-r isk AI system, of at least six

− months, unless provid ed other wise in applicable Uni on or national law , in par ticular in Union law on the protect ion of

− personal data.

− Deplo y ers that are financial institutions subject to requirements regard ing their inter nal gove r nance, ar rang ements or

− processes under Uni on financial ser vices law shall maintain the logs as par t of the documentation kep t pursuant to the

− relevant Uni on financ ial ser vice law .

− 7. Bef ore putting into ser vice or using a high-r isk AI syste m at the wo rkplace, deplo y ers who are emplo y ers shall inf or m

− wo rkers’ representatives and the affect ed work ers that the y will be subject to the use of the high-r isk AI system. This

− inf or mation shall be provided, where applicable, in accordance with the r ules and procedures laid do wn in Union and

− national law and practice on inf or mation of work ers and their representatives.

− 8. Deplo y ers of high-r isk AI systems that are public author ities, or Uni on institutions, bodies, offices or age ncies shall

− comply with the registration obliga tions refe r red to in Ar ticle 49. When suc h deplo y ers find that the high-r isk AI system

− that they en visage using has not been registered in the EU database refer red t o in Ar ticle 71, they shall not use that system

− and shall inf or m the provider or the distr ibutor .

− 9. Where applicable, deplo y ers of high-r isk AI systems shall use the inf or mation provid ed under Ar ticle 13 of this

− Regulation to comp ly with their obligation to car r y out a data protection imp act assessment under Ar ticle 35 of Regulation

− (EU) 2016/679 or Ar ticle 27 of Directive (EU) 2016/680.

− 10. Without prejudice to Directive (EU) 2016/680, in the framew ork of an in vestigation f or the targe te d search of

− a person suspect ed or convict ed of hav ing committed a cr iminal offe nce, the deplo y er of a high-r isk AI syste m f or

− post-remote biometr ic identifica tion shall request an author isation, ex ante, or without undue dela y and no later than 48

− hours, by a judicial author ity or an administrative author ity whose decision is binding and subject to judicial review , f or the

− use of that syste m, excep t when it is used f or the initial identifica tion of a potential suspect based on objective and ver ifi able

− f acts directly linked to the offe nce. Each use shall be limited to what is str ictly necessar y f or the invest igation of a specific

− cr iminal offence.

− If the author isation requested pursuant to the first subparagraph is rejected, the use of the post-remote biometr ic

− identifica tion syste m linked to that requested author isation shall be st opped with immediate eff ect and the personal data

− linked to the use of the high-r isk AI system f or which the author isation was request ed shall be delete d.

− 68/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− In no case shall suc h high-r isk AI syste m f or post-remote biometr ic identifica tion be used f or law enf orcement pur poses in

− an untarget ed wa y , without any link to a cr iminal offe nce, a cr iminal proceeding, a g enuine and present or genuine and

− f oreseeable threat of a cr iminal offenc e, or the searc h f or a specif ic missing person. It shall be ensured that no decision that

− produces an adverse leg al effect on a person ma y be taken by the law enf orcement author ities based solely on the output of

− suc h post-remote biometr ic identifi cation syste ms.

− This paragraph is without prejudice to Ar ticle 9 of Regulation (EU) 2016/679 and Ar ticle 10 of Directive (EU) 2016/680

− f or the processing of biometr ic data.

− Reg ardless of the pur pose or deplo y er , each use of such high-r isk AI syste ms shall be documented in the relevant police file

− and shall be made a vailable t o the relevant marke t sur veillance author ity and the national data protection author ity upon

− request, ex cluding the disclosure of sensitive operational data related to law enf orcement. This subparagraph shall be

− without prejudice to the po wers confe r red b y Directive (EU) 2016/680 on super visor y author ities.

− Deplo y ers shall submit annual repor ts to the relevant marke t sur veillance and national data prot ection author ities on their

− use of post-remote biometr ic identifica tion systems, ex cluding the disclosure of sensitive operational data related to la w

− enf orcement. The repor ts ma y be aggrega ted to cover more than one deplo yment.

− Member States ma y introduce, in accordance with Uni on law , more restr ictive laws on the use of post-remote biometr ic

− identifica tion systems.

− 11. Without prejudice to Ar ticle 50 of this Regulation, deplo y ers of high-r isk AI systems refe r red t o in Annex III that

− make decisions or assist in making decisions relate d to natural persons shall inf or m the natural persons that the y are subject

− to the use of the high-r isk AI syste m. For high-r isk AI systems used f or law enf orcement pur poses Ar ticle 13 of Directive

− (EU) 2016/680 shall apply .

− 12. Deplo yers shall cooperate with the relevant compet ent author ities in any action those author ities take in relation to

− the high-r isk AI system in order to imp lement this Regulation.

− Ar ticle 27

− F undament al r ights impact assessment f or high-r isk AI sys tems

− 1. Pr ior to deplo ying a high-r isk AI syste m refe r red to in Ar ticle 6(2), with the excep tion of high-r isk AI syste ms

− intende d to be used in the area listed in point 2 of Annex III, deplo y ers that are bodies gove r ned b y public law , or are pr ivate

− entities provid ing public ser vices, and deplo y ers of high-r isk AI syste ms refer red t o in points 5 (b) and (c) of Annex III, shall

− perfo r m an assessment of the im pact on fundamental r ights that the use of such syste m ma y produce. For that pur pose,

− deplo yers shall perfo r m an assessment consisting of:

− (a) a descr iption of the deplo y er ’s processes in which the high-r isk AI system will be used in line with its intende d pur pose;

− (b) a descr iption of the per iod of time within which, and the frequency with which, each high-r isk AI system is intended to

− be used;

− (c) the cate gor ies of natural persons and groups likely to be affe cted b y its use in the specif ic cont ext ;

− (d) the specif ic r isk s of har m likely t o hav e an imp act on the categori es of natural persons or groups of persons identifie d

− pursuant to point (c) of this paragraph, taking into account the inf or mation given by the provider pursuant to

− Ar ticle 13;

− (e) a descr ipt ion of the imp lementation of human oversight measures, according to the instr uctions f or use;

− (f) the measures to be tak en in the case of the materialis ation of those r isk s, including the ar rang ements f or inte r nal

− governance and comp laint mec hanisms.

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 69/144

− 2. The obligation laid do wn in paragraph 1 applies to the fi rst use of the high-r isk AI system. The deplo y er ma y , in

− similar cases, rely on previously conducted fundamental r ights imp act assessments or existing impact assessments car r ied

− out by provid er . If, dur ing the use of the high-r isk AI system, the deplo y er considers that any of the elements listed in

− paragraph 1 has ch anged or is no longe r up to date, the deplo y er shall take the necessar y st eps to update the inf or mation.

− 3. Once the assessment refer red t o in paragraph 1 of this Ar ticle has been perf or med, the deplo y er shall notify the

− marke t sur veillance author ity of its results, submitting the filled- out tem plate refer red to in paragraph 5 of this Ar ticle as

− par t of the notification. In the case refer red to in Ar ticle 46(1), deplo y ers ma y be exem pt from that obligation t o notify .

− 4. If an y of the obliga tions laid do wn in this Ar ticle is already met through the data protection impact assessment

− conducted pursuant t o Ar ticle 35 of Regulation (EU) 2016/679 or Ar ticle 27 of Directive (EU) 2016/680, the fundamental

− r ights imp act assessment refe r red to in paragraph 1 of this Ar ticle shall comp lement that data protection im pact

− assessment.

− 5. The AI Office shall develop a te mplat e f or a questionnaire, including through an auto mated t ool, to facilitat e deplo yers

− in com plying with their obliga tions under this Ar ticle in a simplified manner .

− SECTION 4

− Notifying author ities and notified bodies

− Ar ticle 28

− Notifying author ities

− 1. Each Member Stat e shall designate or establish at least one notifying author ity responsible f or setting up and car r ying

− out the necessar y procedures f or the assessment, designation and notif ication of conf or mity assessment bodies and f or their

− monitoring. Those procedures shall be developed in cooperation between the notifying author ities of all Member States.

− 2. Member States ma y decide that the assessment and monitoring referred to in paragraph 1 is to be car r ied out by

− a national accreditation body within the meaning of, and in accordance with, Regulation (EC) No 765/2008.

− 3. Notifying author ities shall be established, org anised and operated in suc h a wa y that no conf lict of interest ar ises with

− conf or mity assessment bodies, and that the objectivity and impar tiality of their activities are saf eguarded.

− 4. Notifying author ities shall be organi sed in suc h a wa y that decisions relating to the notification of conf or mity

− assessment bodies are taken by compet ent persons diffe rent from those who car r ied out the assessment of those bodies.

− 5. Notifying author ities shall offer or provide neither any activities that conf or mity assessment bodies perform , nor any

− consultancy ser vices on a commercial or competitive basis.

− 6. Notifying author ities shall saf eguard the confidentiality of the inf or mation that they obtain, in accordance with

− Ar ticle 78.

− 7. Notifying author ities shall have an adequate number of com petent personnel at their disposal f or the proper

− perfo r mance of their tasks. Comp etent personnel shall hav e the necessar y exper tise, where applicable, f or their function, in

− fie lds suc h as inf or mation te chnologi es, AI and la w , including the super vision of fundamental r ights.

− Ar ticle 29

− Application of a confor mity assessment body for notif ication

− 1. Conf or mity assessment bodies shall submit an application f or notification to the notifying author ity of the Member

− Stat e in which the y are established.

− 70/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− 2. The application f or notif ication shall be accompanied b y a descr iption of the conf or mity assessment activities, the

− conf or mity assessment module or modules and the types of AI syste ms f or which the conf or mity assessment body claims

− to be compet ent, as well as by an accreditation cer tificate, where one exists, issued by a national accreditation body atte sting

− that the conf or mity assessment body fulfils the requirements laid do wn in Ar ticle 31.

− Any valid document related to existing designations of the applicant notified body under any other Union har monisation

− legislation shall be added.

− 3. Where the conf or mity assessment body concer ned cannot provide an accreditation cer tif icate, it shall provid e the

− notifying author ity with all the documentar y evidence necessar y f or the ver ifi cation, recognition and regular monitoring of

− its comp liance with the requirements laid do wn in Ar ticle 31.

− 4. For notif ied bodies which are designate d under any other Union har monisation legislation, all documents and

− cer tif icat es linked to those designations ma y be used to suppor t their designation procedure under this Regulation, as

− appropr iate. The notified body shall update the documentation refer red to in paragraphs 2 and 3 of this Ar ticle whenever

− relevant ch anges occur , in order to enable the author ity responsible f or notified bodies to monitor and ver ify continuous

− complia nce with all the requirements laid down in Ar ticle 31.

− Ar ticle 30

− Notif ication procedure

− 1. Notifying author ities ma y notify only conf or mity assessment bodies which have satisfied the requirements laid do wn

− in Ar ticle 31.

− 2. Notifying author ities shall notify the Commission and the other Member Stat es, using the electronic notification to ol

− developed and managed by the Commission, of each conf or mity assessment body refer red to in paragraph 1.

− 3. The notification refer red t o in paragraph 2 of this Ar ticle shall include full details of the conf or mity assessment

− activities, the conf or mity assessment module or modules, the types of AI syste ms concer ned, and the relevant attestation of

− compet ence. Where a notification is not based on an accreditation cer tif icate as refer red to in Ar ticle 29(2), the notifying

− author ity shall provide the Commission and the other Member States with documentar y evidence which atte sts to the

− compet ence of the conf or mity assessment body and t o the ar rang ements in place t o ensure that that body will be

− monitor ed regularly and will continue t o satisfy the requirements laid do wn in Ar ticle 31.

− 4. The conf or mity assessment body concer ned ma y per f or m the activities of a notified body only where no objections

− are raised b y the Commission or the other Member Stat es within tw o weeks of a notification by a notifying author ity where

− it includes an accreditation cer tif icate refer red to in Ar ticle 29(2), or within tw o months of a notification by the notifying

− author ity where it includes documentar y evidence refer red to in Ar ticle 29(3).

− 5. Where objections are raised, the Commission shall, without dela y , ent er into consultations with the relevant Member

− Stat es and the conf or mity assessment body . In view thereof, the Commission shall decide whether the author isation is

− justified. The Commission shall address its decision to the Member State concer ned and to the relevant conf or mity

− assessment body .

− Ar ticle 31

− Requirements relating t o notif ied bodies

− 1. A notif ied body shall be established under the national law of a Member Stat e and shall hav e legal personality .

− 2. Notifi ed bodies shall satisfy the org anisational, quality managem ent, resources and process requirements that are

− necessar y t o fulf il their tasks, as well as suitable cybersecur ity requirements.

− 3. The org anisational str ucture, allocation of responsibilities, repor ting lines and operation of notif ied bodies shall

− ensure confi dence in their perf or mance, and in the results of the conf or mity assessment activities that the notif ied bodies

− conduct.

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 71/144

− 4. Notifi ed bodies shall be independent of the provid er of a high-r isk AI syste m in relation to which they per f or m

− conf or mity assessment activities. Notifi ed bodies shall also be independent of any other operat or having an economic

− intere st in high-r isk AI syste ms assessed, as well as of any competit ors of the provider . This shall not preclude the use of

− assessed high-r isk AI syste ms that are necessar y f or the operations of the conf or mity assessment body , or the use of such

− high-r isk AI syste ms f or personal pur poses.

− 5. Neither a conf or mity assessment body , its top-level managem ent nor the personnel responsible f or car r ying out its

− conf or mity assessment tasks shall be directly in volved in the design, development, marke ting or use of high-r isk AI systems,

− nor shall they represent the par ties engag ed in those activities. They shall not enga ge in any activity that might conf lict with

− their independence of judgement or inte gr ity in relation to conf or mity assessment activities f or which they are notified.

− This shall, in par ticular , apply to consultancy ser vices.

− 6. Notifi ed bodies shall be org anised and operat ed so as to saf eguard the independence, objectivity and imp ar tiality of

− their activities. Notifi ed bodies shall document and im plement a str ucture and procedures to safeguard im par tiality and to

− promot e and apply the pr inciples of imp ar tiality throughout their org anisation, personnel and assessment activities.

− 7. Notifi ed bodies shall have documented procedures in place ensur ing that their personnel, committ ees, subsidiar ies,

− subcontractors and any associated body or personnel of exte r nal bodies maintain, in accordance with Ar ticle 78, the

− confi dentiality of the inf or mation which comes into their possession dur ing the perf or mance of conf or mity assessment

− activities, ex cept when its disclosure is required by law . The staff of notified bodies shall be bound to obser ve profe ssional

− secrecy with regard t o all inf or mation obtained in car r ying out their tasks under this Regulation, excep t in relation to the

− notifying author ities of the Member Stat e in which their activities are car r ied out.

− 8. Notifi ed bodies shall have procedures f or the perf or mance of activities which take due account of the size of

− a provid er , the sector in which it operat es, its str ucture, and the degree of comp lexity of the AI system concer ned.

− 9. Notifi ed bodies shall tak e out appropr iat e liability insurance f or their conf or mity assessment activities, unless liability

− is assumed b y the Member Stat e in which they are established in accordance with national law or that Member Stat e is itself

− directly responsible f or the conf or mity assessment.

− 10. Notifi ed bodies shall be capable of car r ying out all their tasks under this Regulation with the highest degree of

− profe ssional inte gr ity and the requisite comp etence in the specif ic fie ld, whether those tasks are car r ied out by notif ied

− bodies themselves or on their behalf and under their responsibility .

− 11. Notifi ed bodies shall hav e sufficient inte r nal compet ences to be able effe ctively to evaluate the tasks conducted by

− external par ties on their behalf. The notif ied body shall hav e per manent ava ilability of sufficient administrative, te chnical,

− lega l and scientific personnel who possess exper ience and knowledg e relating t o the relevant types of AI syste ms, data and

− data computing , and relating to the requirements set out in Section 2.

− 12. Notifi ed bodies shall par ticipate in coordination activities as refer red to in Ar ticle 38. They shall also take par t

− directly , or be represente d in, European standardisation organi sations, or ensure that they are aw are and up to dat e in

− respect of relevant standards.

− Ar ticle 32

− Presumption of confor mity with requirements relating to notif ied bodies

− Where a conf or mity assessment body demonstrates its conf or mity with the cr iter ia laid do wn in the relevant har monised

− standards or par ts thereof, the refe rences of which hav e been published in the Off icial Jour nal of the European Union, it shall

− be presumed to comply with the requirements set out in Ar ticle 31 in so f ar as the applicable har monised standards cover

− those requirements.

− 72/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− Ar ticle 33

− Subsidiar ies of notif ied bodies and subcontracting

− 1. Where a notif ied body subcontracts specif ic tasks connect ed with the conf or mity assessment or has recourse to

− a subsidiar y , it shall ensure that the subcontractor or the subsidiar y meets the requirements laid do wn in Ar ticle 31, and

− shall inf or m the notifying author ity according ly .

− 2. Notifi ed bodies shall take full responsibility f or the tasks perf or med by any subcontract ors or subsidiar ies.

− 3. A ctivities ma y be subcontract ed or car r ied out by a subsidiar y only with the agreement of the provider . Notified

− bodies shall mak e a list of their subsidiar ies publicly available.

− 4. The relevant documents concer ning the assessment of the qualific ations of the subcontract or or the subsidiar y and

− the w ork car r ied out by them under this Regulation shall be kep t at the disposal of the notifying author ity f or a per iod of

− five y ears from the t er mination dat e of the subcontracting.

− Ar ticle 34

− Operational obligations of notif ied bodies

− 1. Notifi ed bodies shall ver ify the conf or mity of high-r isk AI systems in accordance with the conf or mity assessment

− procedures set out in Ar ticle 43.

− 2. Notifi ed bodies shall av oid unnecessar y burdens f or pro viders when perf or ming their activities, and take due account

− of the size of the provid er , the secto r in which it operat es, its str ucture and the degree of complexity of the high-r isk AI

− syste m concer ned, in par ticular in view of minimising administrative burdens and compliance costs f or micro- and small

− ent er pr ises within the meaning of Recommendation 2003/361/EC. The notif ied body shall, never theless, respect the degree

− of r igour and the level of prot ection required f or the comp liance of the high-r isk AI system with the requirements of this

− Regulation.

− 3. Notifi ed bodies shall make ava ilable and submit upon request all relevant documentation, including the provider s ’

− documentation, to the notifying author ity refe r red to in Ar ticle 28 to allo w that author ity to conduct its assessment,

− designation, notification and monitoring activities, and to f acilitate the assessment outlined in this Section.

− Ar ticle 35

− Identif ication numbers and lists of notif ied bodies

− 1. The Commission shall assign a sing le identification number to each notif ied body , even where a body is notified under

− more than one Uni on act.

− 2. The Commission shall mak e publicly available the list of the bodies notif ied under this Regulation, including their

− identifica tion numbers and the activities f or which they have been notif ied. The Commission shall ensure that the list is k ep t

− up t o date.

− Ar ticle 36

− Changes to notif ications

− 1. The notifying author ity shall notify the Commission and the other Member States of an y relevant ch anges t o the

− notif ication of a notif ied body via the electronic notification to ol refer red to in Ar ticle 30(2).

− 2. The procedures laid do wn in Ar ticles 29 and 30 shall apply to extensi ons of the scope of the notif ication.

− For ch anges to the notif ication other than extensi ons of its scope, the procedures laid do wn in paragraphs (3) to (9) shall

− apply .

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 73/144

− 3. Where a notif ied body decides to cease its conf or mity assessment activities, it shall inf or m the notifying author ity and

− the provid ers concer ned as soon as possible and, in the case of a planned cessation, at least one y ear bef ore ceasing its

− activities. The cer tif icates of the notified body ma y remain valid f or a per iod of nine months af ter cessation of the notified

− body’s activities, on condition that another notified body has confi r med in wr iting that it will assume responsibilities f or the

− high-r isk AI syste ms covered by those cer tif icat es. The latte r notified body shall complet e a full assessment of the high-r isk

− AI syste ms affect ed by the end of that nine-month-per iod bef ore issuing new cer tif icat es f or those systems. Where the

− notif ied body has ceased its activity , the notifying author ity shall withdra w the designation.

− 4. Where a notifying author ity has suffici ent reason to consider that a notified body no longer meets the requirements

− laid down in Ar ticle 31, or that it is f ailing to fulf il its obliga tions, the notifying author ity shall without dela y in vestig ate the

− matt er with the utmost diligence. In that context, it shall inf or m the notif ied body concer ned about the objections raised

− and give it the possibility to mak e its views known. If the notifying author ity comes to the conclusion that the notified body

− no longer meets the requirements laid do wn in Ar ticle 31 or that it is f ailing to fulfil its obligations, it shall restr ict, suspend

− or withdra w the designation as appropr iat e, depending on the ser iousness of the f ailure to meet those requirements or fulf il

− those obliga tions. It shall immediately inf or m the Commission and the other Member States according ly .

− 5. Where its designation has been suspended, restr icted, or fully or par tially withdra wn, the notif ied body shall inf or m

− the provid ers concer ned within 10 da ys.

− 6. In the event of the restr iction, suspension or withdrawal of a designation, the notifying author ity shall take

− appropr iate steps to ensure that the fi les of the notif ied body concer ned are k ept, and to make them a vailable t o notifying

− author ities in other Member States and to mark et sur veillance author ities at their request.

− 7. In the event of the restr iction, suspension or withdra wal of a designation, the notifying author ity shall:

− (a) assess the imp act on the cer tif icates issued b y the notif ied body ;

− (b) submit a repor t on its fi ndings to the Commission and the other Member Stat es within three months of hav ing notified

− the ch ange s to the designation;

− (c) require the notif ied body to suspend or withdra w , within a reasonable per iod of time det er mined by the author ity , any

− cer tif icat es which were unduly issued, in order to ensure the continuing conf or mity of high-r isk AI systems on the

− marke t ;

− (d) inf or m the Commission and the Member States about cer tif icat es the suspension or withdra wal of which it has required;

− (e) provid e the national compet ent author ities of the Member Stat e in which the provider has its regist ered place of

− business with all relevant inf or mation about the cer tificates of which it has required the suspension or withdra wal; that

− author ity shall take the appropr iate measures, where necessar y , t o av oid a pote ntial r isk to health, safety or fundamental

− r ights.

− 8. With the ex ception of cer tif icat es unduly issued, and where a designation has been suspended or restr icted, the

− cer tif icat es shall remain valid in one of the f ollowing circumstances:

− (a) the notifying author ity has conf ir med, within one month of the suspension or restr iction, that there is no r isk to health,

− saf ety or fundamental r ights in relation to cer tif icat es affect ed by the suspension or restr iction, and the notifying

− author ity has outlined a timeline f or actions to remedy the suspension or restr iction; or

− (b) the notifying author ity has confir med that no cer tif icat es relevant to the suspension will be issued, amended or re-issued

− dur ing the course of the suspension or restr iction, and states whether the notif ied body has the capability of continuing

− to monitor and remain responsible f or existing cer tif icat es issued f or the per iod of the suspension or restr iction; in the

− event that the notifying author ity determ ines that the notif ied body does not have the capability to suppor t existing

− cer tif icat es issued, the provider of the system cover ed by the cer tif icate shall conf ir m in wr iting t o the national

− compet ent author ities of the Member Stat e in which it has its registered place of business, within three months of the

− suspension or restr iction, that another qualif ied notif ied body is te mporar ily assuming the functions of the notif ied

− body to monitor and remain responsible f or the cer tif icates dur ing the per iod of suspension or restr iction.

− 74/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− 9. With the excep tion of cer tif icates unduly issued, and where a designation has been withdra wn, the cer tif icat es shall

− remain valid f or a per iod of nine months under the f ollo wing circumstances:

− (a) the national compet ent author ity of the Member State in which the provider of the high-r isk AI syste m cover ed by the

− cer tif icat e has its registered place of business has confi r med that there is no r isk to health, safety or fundamental r ights

− associate d with the high-r isk AI systems concer ned; and

− (b) another notified body has confir med in wr iting that it will assume immediate responsibility f or those AI syste ms and

− complet es its assessment within 12 months of the withdra wal of the designation.

− In the circumstances refe r red to in the fir st subparagraph, the national comp etent author ity of the Member Stat e in which

− the provid er of the syste m covered by the cer tificate has its place of business ma y exte nd the provisional validity of the

− cer tif icat es f or additional per iods of three months, which shall not exceed 12 months in total.

− The national compet ent author ity or the notif ied body assuming the functions of the notified body affect ed by the chang e

− of designation shall immediate ly inf or m the Commission, the other Member States and the other notif ied bodies thereof.

− Ar ticle 37

− Challenge to the competence of notif ied bodies

− 1. The Commission shall, where necessar y , in vestigat e all cases where there are reasons to doubt the comp et ence of

− a notified body or the continued fulfilment by a notif ied body of the requirements laid down in Ar ticle 31 and of its

− applicable responsibilities.

− 2. The notifying author ity shall provid e the Commission, on request, with all relevant inf or mation relating to the

− notif ication or the maintenance of the comp etence of the notified body concer ned.

− 3. The Commission shall ensure that all sensitive inf or mation obtained in the course of its in vestigations pursuant t o this

− Ar ticle is treate d confidentially in accordance with Ar ticle 78.

− 4. Where the Commission ascer tains that a notif ied body does not meet or no longer meets the requirements f or its

− notif ication, it shall inf or m the notifying Member Stat e according ly and request it to take the necessar y cor rective measures,

− including the suspension or withdra wal of the notif ication if necessar y . Where the Member Stat e f ails to take the necessar y

− cor rective measures, the Commission ma y , by means of an imp lementing act, suspend, restr ict or withdraw the designation.

− That im plementing act shall be adop te d in accordance with the examination procedure refer red to in Ar ticle 98(2).

− Ar ticle 38

− Coordination of notif ied bodies

− 1. The Commission shall ensure that, with regard to high-r isk AI systems, appropr iate coordination and cooperation

− between notified bodies active in the conf or mity assessment procedures pursuant to this Regulation are put in place and

− properly operat ed in the f or m of a sect oral group of notified bodies.

− 2. Each notifying author ity shall ensure that the bodies notified by it par ticipate in the wo rk of a group refer red to in

− paragraph 1, directly or through designated representatives.

− 3. The Commission shall provid e f or the ex ch ange of kno wledge and best practices between notifying author ities.

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 75/144

− Ar ticle 39

− Conf or mity assessment bodies of third countr ies

− Conf or mity assessment bodies established under the law of a third countr y with which the Union has concluded an

− agreement ma y be author ised to car r y out the activities of notified bodies under this Regulation, provided that they meet

− the requirements laid do wn in Ar ticle 31 or they ensure an equivalent level of compliance.

− SECTION 5

− Standar ds, conf or mity assessment, cer tificates, r egistr ation

− Ar ticle 40

− Har monised s tandards and s tandardisation deliv erables

− 1. High-r isk AI syste ms or g eneral-pur pose AI models which are in conf or mity with har monised standards or par ts

− thereof the references of which hav e been published in the Off icial Jour nal of the Eur opean Union in accordance with

− Regulation (EU) No 1025/2012 shall be presumed to be in conf or mity with the requirements set out in Section 2 of this

− Chapt er or , as applicable, with the obliga tions set out in of Chapt er V , Sections 2 and 3, of this Regulation, to the extent that

− those standards cover those requirements or obliga tions.

− 2. In accordance with Ar ticle 10 of Regulation (EU) No 1025/2012, the Commission shall issue, without undue dela y ,

− standardisation requests coveri ng all requirements set out in Section 2 of this Chap ter and, as applicable, standardisation

− requests coverin g obliga tions set out in Chapt er V , Sections 2 and 3, of this Regulation. The standardisation request shall

− also ask f or deliverables on repor ting and documentation processes to impro ve AI systems’ resource perf or mance, such as

− reducing the high-r isk AI system’s consump tion of energy and of other resources dur ing its lifecy cle, and on the

− energy-eff icient development of g eneral-pur pose AI models. When prepar ing a standardisation request, the Commission

− shall consult the Board and relevant stakeh olders, including the advisor y f or um.

− When issuing a standardisation request t o European standardisation organisations, the Commission shall specify that

− standards hav e to be clear , consiste nt, including with the standards developed in the var ious sect ors f or products covered by

− the existing Uni on har monisation legislation listed in Annex I, and aiming to ensure that high-r isk AI syste ms or

− g eneral-pur pose AI models placed on the marke t or put into ser vice in the Union meet the relevant requirements or

− obliga tions laid do wn in this Regulation.

− The Commission shall request the European standardisation organisations to provide evidence of their best eff or ts to fulf il

− the objectives refer red to in the first and the second subparagraph of this paragraph in accordance with Ar ticle 24 of

− Regulation (EU) No 1025/2012.

− 3. The par ticipants in the standardisation process shall seek to promot e invest ment and inno vation in AI, including

− through increasing legal cer tainty , as well as the comp etitiveness and gro wth of the Uni on marke t, to contr ibut e to

− strengthening g lobal cooperation on standardisation and taking into account existing intern ational standards in the field of

− AI that are consiste nt with Union values, fundamental r ights and interests, and to enhance multi-stake holder gove r nance

− ensur ing a balanced representation of inte rests and the eff ective par ticipation of all relevant stak eholders in accordance with

− Ar ticles 5, 6, and 7 of Regulation (EU) No 1025/2012.

− Ar ticle 41

− Common specif ications

− 1. The Commission ma y adopt, imp lementing acts establishing common specifications f or the requirements set out in

− Section 2 of this Chapt er or , as applicable, f or the obligations set out in Sections 2 and 3 of Chapt er V where the f ollowing

− conditions hav e been fulf illed:

− (a) the Commission has requested, pursuant to Ar ticle 10(1) of Regulation (EU) No 1025/2012, one or more European

− standardisation organisations to draf t a har monised standard f or the requirements set out in Section 2 of this Chapt er ,

− or , as applicable, f or the obliga tions set out in Sections 2 and 3 of Chap ter V , and:

− (i) the request has not been accept ed b y an y of the European standardisation org anisations; or

− 76/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− (ii) the har monised standards addressing that request are not delivered within the deadline set in accordance with

− Ar ticle 10(1) of Regulation (EU) No 1025/2012; or

− (iii) the relevant har monised standards insuff iciently address fundamental r ights concer ns; or

− (iv) the har monised standards do not com ply with the request ; and

− (b) no refere nce to har monised standards covering the requirements refer red to in Section 2 of this Chapt er or , as

− applicable, the oblig ations refe r red to in Sections 2 and 3 of Chap ter V has been published in the Off icial Jour nal of the

− Europe an Union in accordance with Regulation (EU) No 1025/2012, and no suc h reference is expect ed to be published

− within a reasonable per iod.

− When draf ting the common specif ications, the Commission shall consult the advisor y f or um refer red to in Ar ticle 67.

− The implementing acts refe r red t o in the fi rst subparagraph of this paragraph shall be adop ted in accordance with the

− examination procedure refe r red t o in Ar ticle 98(2).

− 2. Bef ore prepar ing a draf t implementing act, the Commission shall inf or m the committee referred to in Ar ticle 22 of

− Regulation (EU) No 1025/2012 that it considers the conditions laid do wn in paragraph 1 of this Ar ticle to be fulf illed.

− 3. High-r isk AI syste ms or g eneral-pur pose AI models which are in conf or mity with the common specific ations refe r red

− to in paragraph 1, or par ts of those specifications, shall be presumed to be in conf or mity with the requirements set out in

− Section 2 of this Chapt er or , as applicable, to comp ly with the oblig ations refer red to in Sections 2 and 3 of Chapt er V , to

− the exte nt those common specifications cover those requirements or those obligations.

− 4. Where a har monised standard is adop ted b y a European standardisation org anisation and proposed to the

− Commission f or the publication of its reference in the Off icial Jour nal of the Eur opean Union, the Commission shall assess the

− har monised standard in accordance with Regulation (EU) No 1025/2012. When refe rence to a har monised standard is

− published in the Off icial Jour nal of the European Union, the Commission shall repeal the imp lementing acts refer red to in

− paragraph 1, or par ts thereof which cover the same requirements set out in Section 2 of this Chap ter or , as applicable, the

− same obliga tions set out in Sections 2 and 3 of Chap ter V .

− 5. Where provider s of high-r isk AI syste ms or g eneral-pur pose AI models do not comply with the common

− specific ations refe r red to in paragraph 1, they shall duly justify that the y hav e adopt ed tec hnical solutions that meet the

− requirements refer red to in Section 2 of this Chapt er or , as applicable, comply with the obligations set out in Sections 2 and

− 3 of Chapt er V to a level at least equivalent thereto.

− 6. Where a Member Stat e considers that a common specification does not entirely meet the requirements set out in

− Section 2 or , as applicable, com ply with obliga tions set out in Sections 2 and 3 of Chapt er V , it shall inf or m the

− Commission thereof with a detailed explanation. The Commission shall assess that inf or mation and, if appropr iate, amend

− the im plementing act establishing the common specif ication concer ned.

− Ar ticle 42

− Presumption of confor mity with cer tain requirements

− 1. High-r isk AI systems that hav e been trained and test ed on data ref lecting the specif ic g eographical, behavi oural,

− cont extual or functional setting within which they are intended to be used shall be presumed to comply with the relevant

− requirements laid do wn in Ar ticle 10(4).

− 2. High-r isk AI syste ms that hav e been cer tif ied or f or which a stat ement of conf or mity has been issued under

− a cybersecur ity scheme pursuant to Regulation (EU) 2019/881 and the references of which hav e been published in the

− Off icial Jour nal of the European Union shall be presumed t o comply with the cybersecur ity requirements set out in Ar ticle 15

− of this Regulation in so f ar as the cybersecur ity cer tif icate or statem ent of conf or mity or par ts thereof cover those

− requirements.

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 77/144

− Ar ticle 43

− Conformit y assessment

− 1. For high-r isk AI systems listed in point 1 of Annex III, where, in demonstrating the complia nce of a high-r isk AI

− syste m with the requirements set out in Section 2, the provid er has applied har monised standards refer red t o in Ar ticle 40,

− or , where applicable, common specifications refer red to in Ar ticle 41, the provider shall opt f or one of the f ollowing

− conf or mity assessment procedures based on:

− (a) the inte r nal control refer red to in Annex VI; or

− (b) the assessment of the quality management syste m and the assessment of the te ch nical documentation, with the

− inv olvement of a notified body , refe r red t o in Annex VII.

− In demonstrating the compliance of a high-r isk AI system with the requirements set out in Section 2, the provid er shall

− f ollo w the conf or mity assessment procedure set out in Annex VII where:

− (a) har monised standards refer red t o in Ar ticle 40 do not exist, and common specifications refe r red to in Ar ticle 41 are not

− ava ilable;

− (b) the provid er has not applied, or has applied only par t of, the har monised standard;

− (c) the common specifications refer red to in point (a) exist, but the provider has not applied them;

− (d) one or more of the har monised standards refe r red to in point (a) has been published with a restr iction, and only on the

− par t of the standard that was restr icted.

− For the pur poses of the conf or mity assessment procedure refer red t o in Annex VII, the provider ma y ch oose an y of the

− notif ied bodies. How ever , where the high-r isk AI syste m is intended to be put into ser vice by law enf orcement, immigration

− or asylum author ities or b y Uni on institutions, bodies, off ices or age ncies, the mark et sur veillance author ity refer red t o in

− Ar ticle 74(8) or (9), as applicable, shall act as a notified body .

− 2. For high-r isk AI systems refer red to in points 2 to 8 of Annex III, provider s shall f ollow the conf or mity assessment

− procedure based on inter nal control as refe r red to in Annex VI, which does not provide f or the inv olvement of a notif ied

− body .

− 3. For high-r isk AI systems cover ed b y the Uni on har monisation legislation listed in Section A of Annex I, the provid er

− shall f ollo w the relevant conf or mity assessment procedure as required under those legal acts. The requirements set out in

− Section 2 of this Chapt er shall apply to those high-r isk AI syste ms and shall be par t of that assessment. P oints 4.3., 4.4., 4.5.

− and the fif th paragraph of point 4.6 of Annex VII shall also apply .

− For the pur poses of that assessment, notif ied bodies which hav e been notif ied under those legal acts shall be entitled to

− control the conf or mity of the high-r isk AI syste ms with the requirements set out in Section 2, pro vided that the compliance

− of those notif ied bodies with requirements laid do wn in Ar ticle 31(4), (5), (10) and (11) has been assessed in the context of

− the notification procedure under those lega l acts.

− Where a legal act listed in Section A of Annex I enables the product manufa cturer to opt out from a third-par ty conf or mity

− assessment, provided that that manufacture r has applied all har monised standards coveri ng all the relevant requirements,

− that manufa cturer ma y use that option only if it has also applied har monised standards or , where applicable, common

− specific ations refe r red to in Ar ticle 41, covering all requirements set out in Section 2 of this Chapt er .

− 4. High-r isk AI systems that hav e already been subject to a conf or mity assessment procedure shall undergo a new

− conf or mity assessment procedure in the event of a substantial modification, regardless of whether the modified syste m is

− intende d to be fur ther distr ibut ed or continues t o be used b y the cur rent deplo y er .

− For high-r isk AI systems that continue t o lear n af te r being placed on the marke t or put into ser vice, chang es to the high-r isk

− AI syste m and its perfo r mance that have been pre-dete r mined b y the provid er at the moment of the initial conf or mity

− assessment and are par t of the inf or mation contained in the tec hnical documentation refer red to in point 2(f) of Annex IV ,

− shall not constitute a substantial modifi cation.

− 5. The Commission is empo wered to adop t deleg ated acts in accordance with Ar ticle 97 in order to amend Annex es VI

− and VII b y updating them in light of tec hnical progress.

− 78/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− 6. The Commission is empo wered to adop t deleg ated acts in accordance with Ar ticle 97 in order to amend paragraphs 1

− and 2 of this Ar ticle in order to subject high-r isk AI syste ms refe r red t o in points 2 to 8 of Annex III to the conf or mity

− assessment procedure refe r red to in Annex VII or par ts thereof. The Commission shall adopt such delegat ed acts taking into

− account the effectiveness of the conf or mity assessment procedure based on inte r nal control refe r red to in Annex VI in

− preventing or minimising the r isks t o health and saf ety and prot ection of fundamental r ights posed by suc h syste ms, as well

− as the ava ilability of adequat e capacities and resources among notified bodies.

− Ar ticle 44

− Cer tif icates

− 1. Cer tif icates issued by notif ied bodies in accordance with Annex VII shall be dra wn-up in a languag e which can be

− easily understood by the relevant author ities in the Member State in which the notified body is established.

− 2. Cer tif icates shall be valid f or the per iod they indicate , which shall not ex ceed five y ears f or AI systems covered by

− Annex I, and f our y ears f or AI syste ms co vered b y Annex III. At the request of the provider , the validity of a cer tif icate ma y

− be exte nded f or fur ther per iods, each not ex ceeding fiv e y ears f or AI systems co vered b y Annex I, and f our years f or AI

− syste ms cover ed b y Annex III, based on a re-assessment in accordance with the applicable conf or mity assessment

− procedures. Any supplement to a cer tif icate shall remain valid, provided that the cer tif icat e which it supplements is valid.

− 3. Where a notified body finds that an AI syste m no longer meets the requirements set out in Section 2, it shall, taking

− account of the pr inciple of propor tionality , suspend or withdra w the cer tif icat e issued or impose restr ictions on it, unless

− complia nce with those requirements is ensured by appropr iat e cor rective action take n by the provider of the system within

− an appropr iate deadline set b y the notif ied body . The notif ied body shall give reasons f or its decision.

− An appeal procedure aga inst decisions of the notified bodies, including on conf or mity cer tif icat es issued, shall be available .

− Ar ticle 45

− Infor mation obligations of notif ied bodies

− 1. Notifi ed bodies shall inf or m the notifying author ity of the f ollo wing:

− (a) any Uni on tec hnical documentation assessment cer tificates, an y supplements to those cer tificates, and any quality

− managem ent syste m approva ls issued in accordance with the requirements of Annex VII;

− (b) any refusal, restr iction, suspension or withdra wal of a Union te chnical documentation assessment cer tif icat e or a quality

− managem ent syste m approva l issued in accordance with the requirements of Annex VII;

− (c) any circumstances affecting the scope of or conditions f or notif ication;

− (d) any request f or inf or mation which the y have received from marke t sur veillance author ities regarding conf or mity

− assessment activities;

− (e) on request, conf or mity assessment activities perf or med within the scope of their notif ication and any other activity

− perfo r med, including cross-border activities and subcontracting.

− 2. Each notif ied body shall inf or m the other notif ied bodies of:

− (a) quality managem ent system approva ls which it has refused, suspended or withdra wn, and, upon request, of quality

− syste m approva ls which it has issued;

− (b) Uni on tec hnical documentation assessment cer tificates or any supplements thereto which it has refused, withdra wn,

− suspended or other wise restr icte d, and, upon request, of the cer tif icates and/or supplements thereto which it has issued.

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 79/144

− 3. Each notif ied body shall provide the other notified bodies car r ying out similar conf or mity assessment activities

− covering the same types of AI systems with relevant inf or mation on issues relating to negative and, on request, positive

− conf or mity assessment results.

− 4. Notifi ed bodies shall safegua rd the confi dentiality of the inf or mation that they obtain, in accordance with Ar ticle 78.

− Ar ticle 46

− Derogation from confor mity assessment procedure

− 1. By wa y of derogat ion from Ar ticle 43 and upon a duly justified request, any market sur veillance author ity ma y

− author ise the placing on the marke t or the putting into ser vice of specific high-r isk AI systems within the te r r itory of the

− Member Stat e concer ned, f or ex ceptiona l reasons of public secur ity or the prot ection of life and health of persons,

− envir onmental protection or the prot ection of k ey industr ial and infrastr uctural assets. That author isation shall be f or

− a limited per iod while the necessar y conf or mity assessment procedures are being car r ied out, taking into account the

− ex ceptional reasons justifying the deroga tion. The completion of those procedures shall be under taken without undue dela y .

− 2. In a duly justifi ed situation of urg ency f or excep tional reasons of public secur ity or in the case of specif ic, substantial

− and imminent threat to the life or physica l safety of natural persons, law -enf orcement author ities or civil prot ection

− author ities ma y put a specif ic high-r isk AI system into ser vice without the author isation refe r red t o in paragraph 1,

− provid ed that suc h author isation is request ed dur ing or af ter the use without undue dela y . If the author isation referred to in

− paragraph 1 is refused, the use of the high-r isk AI system shall be stopp ed with immediate eff ect and all the results and

− outputs of suc h use shall be immediately discarded.

− 3. The author isation refe r red to in paragraph 1 shall be issued only if the marke t sur veillance author ity concludes that

− the high-r isk AI syste m comp lies with the requirements of Section 2. The marke t sur veillance author ity shall inf or m the

− Commission and the other Member Stat es of any author isation issued pursuant to paragraphs 1 and 2. This obligation shall

− not co ver sensitive operational data in relation to the activities of la w-enf orcement author ities.

− 4. Where, within 15 calendar da ys of receipt of the inf or mation refer red to in paragraph 3, no objection has been raised

− by either a Member State or the Commission in respect of an author isation issued by a marke t sur veillance author ity of

− a Member State in accordance with paragraph 1, that author isation shall be deemed justified.

− 5. Where, within 15 calendar da ys of receipt of the notification referred to in paragraph 3, objections are raised by

− a Member Stat e against an author isation issued by a mark et sur veillance author ity of another Member State, or where the

− Commission considers the author isation to be contrar y to Uni on law , or the conclusion of the Member Stat es rega rding the

− complia nce of the syste m as refe r red to in paragraph 3 to be unf ounded, the Commission shall, without dela y , ent er into

− consultations with the relevant Member Stat e. The operat ors concer ned shall be consulted and hav e the possibility to

− present their views. Ha ving regard thereto, the Commission shall decide whether the author isation is justified. The

− Commission shall address its decision to the Member Stat e concer ned and to the relevant operato rs.

− 6. Where the Commission considers the author isation unjustified, it shall be withdra wn b y the marke t sur veillance

− author ity of the Member State concer ned.

− 7. For high-r isk AI syste ms related t o products covered by Union har monisation legislation listed in Section A of

− Annex I, only the derogat ions from the conf or mity assessment established in that Uni on har monisation legislation shall

− apply .

− Ar ticle 47

− EU declaration of confor mity

− 1. The provider shall dra w up a wr itt en machine readable, physical or electronically signed EU declaration of conf or mity

− f or each high-r isk AI system, and k eep it at the disposal of the national comp etent author ities f or 10 y ears af ter the

− high-r isk AI syste m has been placed on the mark et or put into ser vice. The EU declaration of conf or mity shall identify the

− high-r isk AI system f or which it has been drawn up. A cop y of the EU declaration of conf or mity shall be submitt ed t o the

− relevant national comp etent author ities upon request.

− 80/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− 2. The EU declaration of conf or mity shall state that the high-r isk AI system concer ned meets the requirements set out in

− Section 2. The EU declaration of conf or mity shall contain the inf or mation set out in Annex V , and shall be translated into

− a language that can be easily understoo d by the national compet ent author ities of the Member States in which the high-r isk

− AI system is placed on the mark et or made available.

− 3. Where high-r isk AI syste ms are subject t o other Union har monisation legislation which also requires an EU

− declaration of conf or mity , a sing le EU declaration of conf or mity shall be drawn up in respect of all Uni on law applicable to

− the high-r isk AI system. The declaration shall contain all the inf or mation required t o identify the Uni on har monisation

− legislation to which the declaration relates .

− 4. By drawing up the EU declaration of conf or mity , the provid er shall assume responsibility f or comp liance with the

− requirements set out in Section 2. The pro vider shall k eep the EU declaration of conf or mity up-to-dat e as appropr iate.

− 5. The Commission is emp owered t o adopt delegat ed acts in accordance with Ar ticle 97 in order t o amend Annex V by

− updating the cont ent of the EU declaration of conf or mity set out in that Annex, in order to introduce elements that become

− necessar y in light of tec hnical progress.

− Ar ticle 48

− CE marking

− 1. The CE marking shall be subject to the general pr inciples set out in Ar ticle 30 of Regulation (EC) No 765/2008.

− 2. For high-r isk AI syste ms provided digitally , a digital CE marking shall be used, only if it can easily be accessed via the

− interfac e from which that system is accessed or via an easily accessible machi ne-readable code or other electronic means.

− 3. The CE marking shall be aff ix ed visibly , legibly and indelibly f or high-r isk AI syste ms. Where that is not possible or

− not war rant ed on account of the nature of the high-r isk AI syste m, it shall be affixed to the packag ing or to the

− accom panying documentation, as appropr iat e.

− 4. Where applicable, the CE marking shall be f ollowed b y the identifica tion number of the notified body responsible f or

− the conf or mity assessment procedures set out in Ar ticle 43. The identifica tion number of the notif ied body shall be affixed

− by the body itself or , under its instr uctions, b y the provid er or by the provid er ’s author ised representative. The identification

− number shall also be indicated in any promotional material which mentions that the high-r isk AI system fulf ils the

− requirements f or CE marking.

− 5. Where high-r isk AI systems are subject to other Uni on law which also provid es f or the affixing of the CE marking, the

− CE marking shall indicate that the high-r isk AI system also fulf il the requirements of that other law .

− Ar ticle 49

− Regis tration

− 1. Bef ore placing on the marke t or putting into ser vice a high-r isk AI system listed in Annex III, with the excep tion of

− high-r isk AI syste ms refe r red to in point 2 of Annex III, the provid er or , where applicable, the author ised representative

− shall regist er themselves and their syste m in the EU database refe r red to in Ar ticle 71.

− 2. Bef ore placing on the market or putting into ser vice an AI system f or which the pro vider has concluded that it is not

− high-r isk according to Ar ticle 6(3), that provider or , where applicable, the author ised representative shall regist er

− themselves and that system in the EU database refer red to in Ar ticle 71.

− 3. Bef ore putting into ser vice or using a high-r isk AI syste m listed in Annex III, with the excep tion of high-r isk AI

− syste ms listed in point 2 of Annex III, deplo y ers that are public author ities, Uni on institutions, bodies, off ices or agencies or

− persons acting on their behalf shall regist er themselves, select the syste m and register its use in the EU database referred to

− in Ar ticle 71.

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 81/144

− 4. For high-r isk AI systems refer red to in points 1, 6 and 7 of Annex III, in the areas of law enf orcement, migration,

− asylum and border control management, the registration referred t o in paragraphs 1, 2 and 3 of this Ar ticle shall be in

− a secure non-public section of the EU database refer red to in Ar ticle 71 and shall include only the f ollo wing inf or mation, as

− applicable, referred to in:

− (a) Section A, points 1 to 10, of Annex VIII, with the excep tion of points 6, 8 and 9;

− Only the Commission and national author ities refe r red t o in Ar ticle 74(8) shall hav e access to the respective restr icted

− sections of the EU database listed in the first subparagraph of this paragraph.

− 5. High-r isk AI syste ms refer red to in point 2 of Annex III shall be regist ered at national level.

− C HAPTER IV

− TRANSP ARENCY OBLIGA TIONS F OR PR O VIDERS AND DEPLO YERS OF CERT AIN AI SYSTEMS

− Ar ticle 50

− T ransparency obligations f or pro viders and deplo y ers of cer ta in AI sys tems

− 1. Providers shall ensure that AI syste ms intended to interac t directly with natural persons are designed and developed in

− suc h a wa y that the natural persons concer ned are inf or med that they are interacting with an AI system, unless this is

− obvious from the point of view of a natural person who is reasonably well-inf or med, obser vant and circumspect, taking

− into account the circumstances and the cont ext of use. This obligation shall not apply t o AI systems author ised b y law to

− detect , prevent, invest igat e or prosecute cr iminal offe nces, subject t o appropr iate saf eguards f or the r ights and freedoms of

− third par ties, unless those syste ms are ava ilable f or the public to repor t a cr iminal offence.

− 2. Providers of AI syste ms, including g eneral-pur pose AI systems, g enerating synthetic audio, image, video or text

− cont ent, shall ensure that the outputs of the AI system are marke d in a machine-readable f or mat and detectable as

− ar tif icially generat ed or manipulate d. Providers shall ensure their t echnical solutions are effe ctive, inte roperable, robust and

− reliable as far as this is t echnically f easible, taking into account the specif icities and limitations of var ious types of cont ent,

− the costs of imp lementation and the g enerally ackno wledged state of the ar t, as ma y be ref lected in relevant t echnical

− standards. This obligation shall not apply t o the extent the AI systems perfo r m an assistive function f or standard editing or

− do not substantially alt er the input data provid ed by the deplo y er or the semantics thereof, or where author ised b y law to

− detect , prevent, investig ate or prosecute cr iminal offe nces.

− 3. Deplo y ers of an emotion recognition syste m or a biometr ic catego r isation syste m shall inf or m the natural persons

− exposed thereto of the operation of the syste m, and shall process the personal data in accordance with Regulations (EU)

− 2016/679 and (EU) 2018/1725 and Directive (EU) 2016/680, as applicable. This obliga tion shall not apply to AI syste ms

− used f or biometr ic categor isation and emotion recognition, which are per mitted b y law to detect , prevent or invest igat e

− cr iminal offe nces, subject to appropr iate saf eguards f or the r ights and freedoms of third par ties, and in accordance with

− Uni on law .

− 4. Deplo y ers of an AI syste m that gene rates or manipulat es image, audio or video cont ent constituting a deep f ake, shall

− disclose that the content has been ar tificially g enerated or manipulated. This obliga tion shall not apply where the use is

− author ised by law to detect, prevent, invest igat e or prosecute cr iminal offence. Where the cont ent f or ms par t of an evidently

− ar tistic, creative, satir ical, fictional or analogous work or programme, the transparency obligations set out in this paragraph

− are limit ed to disclosure of the existe nce of suc h g enerated or manipulate d cont ent in an appropr iate manner that does not

− ham per the displa y or enjo yment of the work.

− Deplo y ers of an AI system that generat es or manipulate s te xt which is published with the pur pose of inf or ming the public

− on matt ers of public intere st shall disclose that the text has been ar tif icially g enerated or manipulate d. This obligation shall

− not apply where the use is author ised by law to detect, prevent, investig ate or prosecute cr iminal offences or where the

− AI-g enerated cont ent has undergone a process of human review or edito r ial control and where a natural or legal person

− holds editor ial responsibility f or the publication of the content.

− 82/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− 5. The inf or mation refe r red t o in paragraphs 1 to 4 shall be pro vided to the natural persons concer ned in a clear and

− distinguishable manner at the latest at the time of the first inte raction or exposure. The inf or mation shall conf or m t o the

− applicable accessibility requirements.

− 6. Paragraphs 1 to 4 shall not affect the requirements and obligations set out in Chapt er III, and shall be without

− prejudice to other transparency oblig ations laid do wn in Union or national law f or deplo y ers of AI systems.

− 7. The AI Office shall encourage and f acilitate the dra wing up of codes of practice at Uni on level to f acilitate the eff ective

− imp lementation of the obligations regard ing the det ection and labelling of ar tif icially g enerated or manipulat ed cont ent.

− The Commission ma y adop t imp lementing acts to approve those codes of practice in accordance with the procedure laid

− do wn in Ar ticle 56 (6). If it deems the code is not adequat e, the Commission ma y adopt an imp lementing act specifying

− common r ules f or the implementation of those obliga tions in accordance with the examination procedure laid do wn in

− Ar ticle 98(2).

− C HAPTER V

− GENERAL -PURPOSE AI MODELS

− SECTION 1

− Classification r ules

− Ar ticle 51

− Classif ication of general-pur pose AI models as general-pur pose AI models with sys temic r isk

− 1. A g eneral-pur pose AI model shall be classified as a g eneral-pur pose AI model with systemic r isk if it meets any of the

− f ollo wing conditions:

− (a) it has high imp act capabilities evaluated on the basis of appropr iate tec hnical tools and methodologies, including

− indicator s and benchmarks;

− (b) based on a decision of the Commission, ex of ficio or f ollowi ng a qualif ied aler t from the scientific panel, it has

− capabilities or an imp act equivalent to those set out in point (a) hav ing rega rd to the cr iter ia set out in Annex XIII.

− 2. A g eneral-pur pose AI model shall be presumed t o hav e high im pact capabilities pursuant to paragraph 1, point (a),

− when the cumulative amount of computation used f or its training measured in f loating point operations is great er than

− 10

− 25

− .

− 3. The Commission shall adopt delegat ed acts in accordance with Ar ticle 97 t o amend the thresholds listed in

− paragraphs 1 and 2 of this Ar ticle, as well as to supplement benchmarks and indicator s in light of evolving te chnological

− developments, suc h as algor ithmic im provements or increased hardware efficiency , when necessar y , f or these thresholds to

− ref lect the state of the ar t.

− Ar ticle 52

− Procedure

− 1. Where a gene ral-pur pose AI model meets the condition refe r red to in Ar ticle 51(1), point (a), the relevant provid er

− shall notify the Commission without dela y and in any event within two week s af te r that requirement is met or it becomes

− kno wn that it will be met. That notif ication shall include the inf or mation necessar y to demonstrate that the relevant

− requirement has been met. If the Commission becomes aw are of a g eneral-pur pose AI model presenting syste mic r isk s of

− which it has not been notified, it ma y decide to designate it as a model with syste mic r isk.

− 2. The provid er of a general-pur pose AI model that meets the condition refer red t o in Ar ticle 51(1), point (a), ma y

− present, with its notif ication, suff iciently substantiate d arguments to demonstrate that, ex ceptiona lly , although it meets that

− requirement, the g eneral-pur pose AI model does not present, due to its specif ic ch aracteristics , syste mic r isk s and theref ore

− should not be classif ied as a general-pur pose AI model with systemic r isk.

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 83/144

− 3. Where the Commission concludes that the arguments submitted pursuant to paragraph 2 are not suffi ciently

− substantiate d and the relevant provid er was not able to demonstrate that the general-purpo se AI model does not present,

− due t o its specif ic ch aracter istics, syste mic r isks , it shall reject those arguments, and the g eneral-pur pose AI model shall be

− considered to be a g eneral-pur pose AI model with syste mic r isk.

− 4. The Commission ma y designate a g eneral-pur pose AI model as presenting systemic r isks, ex of ficio or f ollowing

− a qualif ied aler t from the scientific panel pursuant to Ar ticle 90(1), point (a), on the basis of cr iter ia set out in Annex XIII.

− The Commission is emp owered to adop t delegat ed acts in accordance with Ar ticle 97 in order to amend Annex XIII by

− specifying and updating the cr iter ia set out in that Annex.

− 5. Upon a reasoned request of a provid er whose model has been designate d as a general-pur pose AI model with systemic

− r isk pursuant to paragraph 4, the Commission shall take the request into account and ma y decide to reassess whether the

− g eneral-pur pose AI model can still be considered to present systemic r isks on the basis of the cr ite r ia set out in Annex XIII.

− Such a request shall contain objective, detailed and new reasons that hav e ar isen since the designation decision. Provi ders

− ma y request reassessment at the earliest six months af t er the designation decision. Where the Commission, f ollowing its

− reassessment, decides to maintain the designation as a general-pur pose AI model with syste mic r isk, providers ma y request

− reassessment at the earliest six months af te r that decision.

− 6. The Commission shall ensure that a list of g eneral-pur pose AI models with systemic r isk is published and shall kee p

− that list up t o dat e, without prejudice to the need to obser ve and prot ect inte llectual proper ty r ights and confi dential

− business inf or mation or trade secrets in accordance with Uni on and national law .

− SECTION 2

− Oblig ations f or providers of g ener al-pur pose AI models

− Ar ticle 53

− Obligations for pro viders of general-pur pose AI models

− 1. Providers of g eneral-pur pose AI models shall:

− (a) dra w up and keep up-to-da te the te chnical documentation of the model, including its training and testing process and

− the results of its evaluation, which shall contain, at a minimum, the inf or mation set out in Annex XI f or the pur pose of

− provid ing it, upon request, to the AI Office and the national compet ent author ities;

− (b) dra w up, keep up-to-da te and mak e a vailable inf or mation and documentation t o provid ers of AI syste ms who inte nd to

− integrat e the g eneral-pur pose AI model into their AI systems. Without prejudice to the need t o obser ve and prot ect

− intellect ual proper ty r ights and confidential business inf or mation or trade secrets in accordance with Union and

− national law , the inf or mation and documentation shall:

− (i) enable provider s of AI syste ms to hav e a good understanding of the capabilities and limitations of the

− g eneral-pur pose AI model and to comply with their obligations pursuant to this Regulation; and

− (ii) contain, at a minimum, the elements set out in Annex XII;

− (c) put in place a policy to comply with Uni on law on cop yr ight and related r ights, and in par ticular to identify and comply

− with, including through state-of-t he-ar t te chnologi es, a reser vation of r ights expressed pursuant to Ar ticle 4(3) of

− Directive (EU) 2019/790;

− (d) dra w up and make publicly availa ble a sufficiently detailed summar y about the content used f or training of the

− g eneral-pur pose AI model, according to a te mplat e provided by the AI Office.

− 84/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− 2. The obliga tions set out in paragraph 1, points (a) and (b), shall not apply to providers of AI models that are released

− under a free and open-source licence that allo ws f or the access, usage , modif ication, and distr ibution of the model, and

− whose paramet ers, including the weights, the inf or mation on the model archit ecture, and the inf or mation on model usage,

− are made publicly available . This ex ception shall not apply to g eneral-pur pose AI models with systemic r isks .

− 3. Providers of g eneral-pur pose AI models shall cooperate as necessar y with the Commission and the national

− compet ent author ities in the ex ercise of their comp etences and powers pursuant to this Regulation.

− 4. Providers of g eneral-pur pose AI models ma y rely on codes of practice within the meaning of Ar ticle 56 to

− demonstrate complia nce with the obliga tions set out in paragraph 1 of this Ar ticle, until a har monised standard is

− published. Compliance with European har monised standards grants providers the presump tion of conf or mity t o the exte nt

− that those standards cover those obliga tions. Providers of g eneral-pur pose AI models who do not adhere to an approved

− code of practice or do not comply with a European har monised standard shall demonstrate alt er native adequat e means of

− complia nce f or assessment by the Commission.

− 5. For the pur pose of facilitating compliance with Annex XI, in par ticular points 2 (d) and (e) thereof, the Commission is

− emp owered to adop t deleg ated acts in accordance with Ar ticle 97 to detail measurement and calculation methodologies

− with a view to allowing f or comp arable and ver ifi able documentation.

− 6. The Commission is empo wered to adop t deleg ated acts in accordance with Ar ticle 97(2) to amend Annexes XI and XII

− in light of ev olving t echnological developments.

− 7. Any inf or mation or documentation obtained pursuant to this Ar ticle, including trade secrets, shall be treate d in

− accordance with the conf identiality obliga tions set out in Ar ticle 78.

− Ar ticle 54

− Author ised represent ativ es of pro viders of general-pur pose AI models

− 1. Pr ior to placing a general-pur pose AI model on the Uni on market, provider s established in third countr ies shall, by

− wr itten mandat e, appoint an author ised representative which is established in the Union.

− 2. The provid er shall enable its author ised representative to perf or m the tasks specified in the mandate received from the

− provid er .

− 3. The author ised representative shall perfo r m the tasks specified in the mandate received from the provider . It shall

− provid e a cop y of the mandate to the AI Offi ce upon request, in one of the off icial languag es of the institutions of the

− Uni on. For the pur poses of this Regulation, the mandate shall empo wer the author ised representative to car r y out the

− f ollo wing tasks :

− (a) ver ify that the tec hnical documentation specif ied in Annex XI has been drawn up and all obliga tions refe r red to in

− Ar ticle 53 and, where applicable, Ar ticle 55 have been fulfilled b y the provider;

− (b) keep a cop y of the te ch nical documentation specified in Annex XI at the disposal of the AI Offi ce and national

− compet ent author ities, f or a per iod of 10 y ears af ter the g eneral-pur pose AI model has been placed on the mark et, and

− the contact details of the provid er that appointed the author ised representative;

− (c) provid e the AI Office, upon a reasoned request, with all the inf or mation and documentation, including that refe r red to

− in point (b), necessar y to demonstrat e complia nce with the oblig ations in this Chapt er;

− (d) cooperate with the AI Office and compet ent author ities, upon a reasoned request, in any action they take in relation to

− the g eneral-pur pose AI model, including when the model is integrat ed into AI systems placed on the marke t or put into

− ser vice in the Uni on.

− 4. The mandate shall em power the author ised representative to be addressed, in addition to or inst ead of the provider ,

− by the AI Offi ce or the compet ent author ities, on all issues related to ensur ing complia nce with this Regulation.

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 85/144

− 5. The author ised representative shall te r minate the mandat e if it considers or has reason to consider the provider to be

− acting contrar y to its obliga tions pursuant to this Regulation. In suc h a case, it shall also immediate ly inf or m the AI Off ice

− about the te r mination of the mandate and the reasons theref or .

− 6. The obliga tion set out in this Ar ticle shall not apply to providers of g eneral-pur pose AI models that are released under

− a free and open-source licence that allows f or the access, usage , modifi cation, and distr ibution of the model, and whose

− parameter s, including the weights, the inf or mation on the model arch itecture, and the inf or mation on model usage , are

− made publicly a vailable, unless the g eneral-pur pose AI models present systemic r isks .

− SECTION 3

− Oblig ations of providers of g ener al-pur pose AI models with syste mic r isk

− Ar ticle 55

− Obligations of pro viders of general-pur pose AI models with sys temic r isk

− 1. In addition t o the obligations listed in Ar ticles 53 and 54, provid ers of g eneral-pur pose AI models with systemic r isk

− shall:

− (a) perfo r m model evaluation in accordance with standardised prot ocols and too ls ref lecting the state of the ar t, including

− conducting and documenting adversar ial te sting of the model with a view to identifying and mitiga ting syste mic r isks ;

− (b) assess and mitig ate possible systemic r isks at Uni on level, including their sources, that ma y st em from the development,

− the placing on the marke t, or the use of general-pur pose AI models with syste mic r isk;

− (c) keep trac k of, document, and repor t, without undue dela y , to the AI Off ice and, as appropr iate, to national compet ent

− author ities, relevant inf or mation about ser ious incidents and possible cor rective measures to address them;

− (d) ensure an adequate level of cybersecur ity protect ion f or the gene ral-pur pose AI model with systemic r isk and the

− physica l infrastr ucture of the model.

− 2. Providers of g eneral-pur pose AI models with systemic r isk ma y rely on codes of practice within the meaning of

− Ar ticle 56 to demonstrate comp liance with the obliga tions set out in paragraph 1 of this Ar ticle, until a har monised

− standard is published. Compliance with European har monised standards grants provid ers the presump tion of conf or mity to

− the extent that those standards cover those obligations. Providers of general-pur pose AI models with syste mic r isks who do

− not adhere to an approved code of practice or do not comply with a European har monised standard shall demonstrate

− alt er native adequate means of compliance f or assessment b y the Commission.

− 3. Any inf or mation or documentation obtained pursuant to this Ar ticle, including trade secrets, shall be treate d in

− accordance with the conf identiality obliga tions set out in Ar ticle 78.

− SECTION 4

− Codes of pr actice

− Ar ticle 56

− Codes of practice

− 1. The AI Offi ce shall encourage and f acilitate the dra wing up of codes of practice at Uni on level in order to contr ibute

− to the proper application of this Regulation, taking into account international approaches.

− 2. The AI Office and the Board shall aim to ensure that the codes of practice co ver at least the obliga tions provid ed f or in

− Ar ticles 53 and 55, including the f ollo wing issues:

− 86/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− (a) the means to ensure that the inf or mation refe r red to in Ar ticle 53(1), points (a) and (b), is kep t up t o dat e in light of

− marke t and te chnological developments;

− (b) the adequat e level of detail f or the summar y about the cont ent used f or training;

− (c) the identifica tion of the type and nature of the systemic r isks at Union level, including their sources, where appropr iate;

− (d) the measures, procedures and modalities f or the assessment and manag ement of the syste mic r isks at Union level,

− including the documentation thereof, which shall be propor tionate t o the r isk s, take into consideration their sever ity

− and probability and take into account the specif ic challeng es of tack ling those r isk s in light of the possible wa ys in

− which suc h r isk s ma y emerge and materi alise along the AI value chain.

− 3. The AI Offi ce ma y in vite all provider s of general-pur pose AI models, as well as relevant national comp etent

− author ities, to par ticipate in the dra wing-up of codes of practice. Civil society organisations, industr y , academia and other

− relevant stak eholders, suc h as do wnstream provider s and independent exper ts, ma y suppor t the process.

− 4. The AI Offi ce and the Board shall aim to ensure that the codes of practice clearly set out their specif ic objectives and

− contain commitments or measures, including ke y perfo r mance indicator s as appropr iate, t o ensure the ac hievement of

− those objectives, and that they take due account of the needs and inte rests of all interest ed par ties, including affect ed

− persons, at Union level.

− 5. The AI Office shall aim to ensure that par ticipants to the codes of practice repor t regularly to the AI Office on the

− imp lementation of the commitments and the measures taken and their outcomes, including as measured agains t the k ey

− perfo r mance indicator s as appropr iate. Ke y perfo r mance indicator s and repor ting commitments shall ref lect diffe rences in

− size and capacity between var ious par ticipants.

− 6. The AI Offi ce and the Board shall regularly monitor and evaluate the achi evement of the objectives of the codes of

− practice b y the par ticipants and their contr ibution to the proper application of this Regulation. The AI Office and the Board

− shall assess whether the codes of practice cover the obliga tions provided f or in Ar ticles 53 and 55, and shall regularly

− monitor and evaluate the achievement of their objectives. They shall publish their assessment of the adequacy of the codes

− of practice.

− The Commission ma y , b y wa y of an implementing act, approve a code of practice and give it a g eneral validity within the

− Uni on. That implementing act shall be adop te d in accordance with the examination procedure refe r red to in Ar ticle 98(2).

− 7. The AI Office ma y invit e all providers of general-pur pose AI models to adhere to the codes of practice. For provider s

− of g eneral-pur pose AI models not presenting syste mic r isks this adherence ma y be limit ed to the obligations provided f or in

− Ar ticle 53, unless the y declare explicitly their interest to join the full code.

− 8. The AI Office shall, as appropr iate, also encourage and f acilitate the review and adaptation of the codes of practice, in

− par ticular in light of emerging standards. The AI Off ice shall assist in the assessment of ava ilable standards.

− 9. Codes of practice shall be ready at the lat est by 2 Ma y 2025. The AI Off ice shall take the necessar y st eps, including

− inviting provider s pursuant to paragraph 7.

− If, by 2 Au gust 2025, a code of practice cannot be final ised, or if the AI Office deems it is not adequat e f ollowing its

− assessment under paragraph 6 of this Ar ticle, the Commission ma y provid e, by means of imp lementing acts, common r ules

− f or the implementation of the obliga tions provided f or in Ar ticles 53 and 55, including the issues set out in paragraph 2 of

− this Ar ticle. Those imp lementing acts shall be adop ted in accordance with the examination procedure refer red to in Ar ticle

− 98(2).

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 87/144

− CHAPTER VI

− MEASURES IN SUPPOR T OF INNO V A TION

− Ar ticle 57

− AI regulator y sandbo xe s

− 1. Member Stat es shall ensure that their compet ent author ities establish at least one AI regulato r y sandbo x at national

− level, which shall be operational by 2 August 2026. That sandbo x ma y also be established jointly with the comp etent

− author ities of other Member States. The Commission ma y provide tec hnical suppor t, advice and to ols f or the establishment

− and operation of AI regulatory sandbo xe s.

− The obligation under the first subparagraph ma y also be fulfilled b y par ticipating in an existing sandbo x in so far as that

− par ticipation provides an equivalent level of national coverag e f or the par ticipating Member Stat es.

− 2. Additional AI regulator y sandbo xes at regional or local level, or established jointly with the compet ent author ities of

− other Member States ma y also be established.

− 3. The European Data Prot ection Super visor ma y also establish an AI regulator y sandbo x f or Uni on institutions, bodies,

− off ices and agencies, and ma y ex ercise the roles and the tasks of national compet ent author ities in accordance with this

− Chapt er .

− 4. Member Stat es shall ensure that the comp etent author ities refe r red to in paragraphs 1 and 2 allocate suffi cient

− resources to comply with this Ar ticle effectively and in a timely manner . Where appropr iate, national compet ent author ities

− shall cooperate with other relevant author ities, and ma y allow f or the in volvement of other actors within the AI ecosyste m.

− This Ar ticle shall not affe ct other regulato r y sandbo xes established under Uni on or national law . Member Stat es shall ensure

− an appropr iate level of cooperation between the author ities super vising those other sandbo xes and the national compet ent

− author ities.

− 5. AI regulatory sandbo xe s established under paragraph 1 shall provid e f or a controlled en vironment that f osters

− inno vation and f acilitates the development, training, te sting and validation of inno vative AI syste ms f or a limited time

− bef ore their being placed on the marke t or put into ser vice pursuant to a specif ic sandbo x plan agreed between the

− provid ers or prospective providers and the compet ent author ity . Such sandbo xes ma y include t esting in real wo rld

− conditions super vised therein.

− 6. Compet ent author ities shall provid e, as appropr iat e, guidance, super vision and suppor t within the AI regulator y

− sandbo x with a view to identifying r isks, in par ticular to fundamental r ights, health and saf ety , t esting, mitigation measures,

− and their effe ctiveness in relation to the obligations and requirements of this Regulation and, where relevant, other Uni on

− and national law super vised within the sandbo x.

− 7. Compet ent author ities shall pro vide provider s and prospective provid ers par ticipating in the AI regulator y sandbo x

− with guidance on regulator y expectations and ho w to fulfil the requirements and obligations set out in this Regulation.

− Upon request of the provider or prospective provid er of the AI system, the compet ent author ity shall provid e a wr itt en

− proof of the activities successfully car r ied out in the sandbo x. The compet ent author ity shall also provide an exit repor t

− detailing the activities car r ied out in the sandbo x and the related results and lear ning outcomes. Provi ders ma y use such

− documentation to demonstrate their compliance with this Regulation through the conf or mity assessment process or

− relevant marke t sur veillance activities. In this regard , the exit repor ts and the wr itten proof provid ed by the national

− compet ent author ity shall be taken positively into account by market sur veillance author ities and notif ied bodies, with

− a view to accelerating conf or mity assessment procedures t o a reasonable exte nt.

− 8. Subject t o the confidentiality provisions in Ar ticle 78, and with the agreement of the provider or prospective provider ,

− the Commission and the Board shall be author ised to access the exit repor ts and shall take them into account, as

− appropr iate, when ex ercising their tasks under this Regulation. If both the provid er or prospective provid er and the national

− compet ent author ity explicitly agree, the exit repor t ma y be made publicly ava ilable through the sing le inf or mation

− platf or m refer red to in this Ar ticle.

− 9. The establishment of AI regulator y sandbo xes shall aim to contr ibute to the f ollowing objectives:

− (a) imp roving legal cer tainty t o achi eve regulator y complia nce with this Regulation or , where relevant, other applicable

− Uni on and national law;

− 88/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− (b) suppor ting the shar ing of best practices through cooperation with the author ities inv olved in the AI regulator y

− sandbo x;

− (c) f oste r ing inno vation and comp etitiveness and f acilitating the development of an AI ecosyste m;

− (d) contr ibuting to evidence-based regulatory lear ning;

− (e) f acilitating and accelerating access to the Union marke t f or AI syste ms, in par ticular when provided by SMEs, including

− star t-ups.

− 10. National comp etent author ities shall ensure that, to the exte nt the innovative AI syste ms inv olve the processing of

− personal data or other wise fall under the super visor y remit of other national author ities or comp etent author ities providing

− or suppor ting access t o data, the national data protect ion author ities and those other national or compet ent author ities are

− associate d with the operation of the AI regulator y sandbo x and in volved in the super vision of those aspects to the extent of

− their respective tasks and powers.

− 11. The AI regulato r y sandbo xes shall not affect the super visor y or cor rective po wers of the compet ent author ities

− super vising the sandbo xes, including at regional or local level. Any significant r isks to health and saf ety and fundamental

− r ights identifi ed dur ing the development and te sting of suc h AI systems shall result in an adequat e mitigation. National

− compet ent author ities shall hav e the po wer to te mporar ily or per manently suspend the te sting process, or the par ticipation

− in the sandbo x if no effective mitig ation is possible, and shall inf or m the AI Offi ce of suc h decision. National compet ent

− author ities shall exer cise their super visor y powe rs within the limits of the relevant law , using their discretionar y po wers

− when imp lementing legal provisions in respect of a specific AI regulator y sandbo x project, with the objective of suppor ting

− inno vation in AI in the Uni on.

− 12. Providers and prospective provid ers par ticipating in the AI regulator y sandbo x shall remain liable under applicable

− Uni on and national liability law f or any damage inf licted on third par ties as a result of the exper imentation taking place in

− the sandbo x. How ever , provid ed that the prospective provider s obser ve the specif ic plan and the t er ms and conditions f or

− their par ticipation and f ollow in good f aith the guidance given by the national compet ent author ity , no administrative fines

− shall be im posed by the author ities f or infr ingements of this Regulation. Where other com petent author ities responsible f or

− other Uni on and national la w were actively inv olved in the super vision of the AI system in the sandbo x and provid ed

− guidance f or compliance, no administrative fine s shall be imp osed rega rding that law .

− 13. The AI regulato r y sandbo xe s shall be designed and imp lemented in suc h a wa y that, where relevant, they facilitat e

− cross-border cooperation between national compet ent author ities.

− 14. National comp etent author ities shall coordinate their activities and cooperate within the framew ork of the Board.

− 15. National comp etent author ities shall inf or m the AI Office and the Board of the establishment of a sandbo x, and ma y

− ask them f or suppor t and guidance. The AI Off ice shall make publicly ava ilable a list of planned and existing sandbo xes and

− keep it up to date in order to encourage more inte raction in the AI regulator y sandbo xes and cross-border cooperation.

− 16. National comp etent author ities shall submit annual repor ts to the AI Office and to the Board, from one year af ter

− the establishment of the AI regulatory sandbo x and ever y y ear thereaf ter until its term ination, and a final repor t. Those

− repor ts shall provid e inf or mation on the progress and results of the imp lementation of those sandbo xes, including best

− practices, incidents, lessons lear nt and recommendations on their setup and, where relevant, on the application and possible

− revision of this Regulation, including its deleg ated and impl ementing acts, and on the application of other Uni on la w

− super vised by the compet ent author ities within the sandbo x. The national compet ent author ities shall make those annual

− repor ts or abstracts thereof a vailable to the public, online. The Commission shall, where appropr iat e, take the annual

− repor ts into account when exer cising its tasks under this Regulation.

− 17. The Commission shall develop a sing le and dedicated interface containing all relevant inf or mation relate d to AI

− regulator y sandbo xe s t o allow stak eholders t o interac t with AI regulato r y sandbo xes and to raise enquir ies with compet ent

− author ities, and to seek non-binding guidance on the conf or mity of inno vative products, ser vices, business models

− embedding AI tec hnologies, in accordance with Ar ticle 62(1), point (c). The Commission shall proactively coordinate with

− national compet ent author ities, where relevant.

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 89/144

− Ar ticle 58

− Detail ed ar rangements f or , and functioning of, AI regulat or y sandbo xes

− 1. In order to av oid fragmentation across the Uni on, the Commission shall adop t imp lementing acts specifying the

− detailed ar rangements f or the establishment, development, im plementation, operation and super vision of the AI regulator y

− sandbo xes. The imp lementing acts shall include common pr inciples on the f ollo wing issues:

− (a) eligibility and selection cr iteria f or par ticipation in the AI regulator y sandbo x;

− (b) procedures f or the application, par ticipation, monitoring, exiting from and te r mination of the AI regulator y sandbo x,

− including the sandbo x plan and the exit repor t ;

− (c) the t er ms and conditions applicable to the par ticipants.

− Those implementing acts shall be adop te d in accordance with the examination procedure refer red to in Ar ticle 98(2).

− 2. The implementing acts referred to in paragraph 1 shall ensure:

− (a) that AI regulator y sandbo xes are open t o an y applying provider or prospective provid er of an AI system who fulfils

− eligibility and selection cr iter ia, which shall be transparent and f air , and that national com petent author ities inf or m

− applicants of their decision within three months of the application;

− (b) that AI regulatory sandbo xes allow broad and equal access and k eep up with demand f or par ticipation; pro viders and

− prospective pro viders ma y also submit applications in par tnerships with deplo y ers and other relevant third par ties;

− (c) that the detailed ar rang ements f or , and conditions concer ning AI regulator y sandbo xe s suppor t, to the best exte nt

− possible, f lexibility f or national comp etent author ities to establish and operate their AI regulatory sandbo xe s;

− (d) that access to the AI regulator y sandbo xes is free of ch arge f or SMEs, including star t-ups, without prejudice to

− ex ceptional costs that national comp etent author ities ma y reco ver in a f air and propor tionate manner ;

− (e) that they facilitat e providers and prospective providers, by means of the lear ning outcomes of the AI regulator y

− sandbo xes, in comp lying with conf or mity assessment obligations under this Regulation and the v oluntar y application

− of the codes of conduct referred to in Ar ticle 95;

− (f) that AI regulator y sandbo xes f acilitate the inv olvement of other relevant actor s within the AI ecosyste m, suc h as notif ied

− bodies and standardisation organisations, SMEs, including star t-ups, ent er pr ises, innovat ors, te sting and exper imenta -

− tion f acilities, research and exper imentation labs and European Digital Inno vation Hubs, centres of ex cellence,

− individual researchers , in order to allow and facilitat e cooperation with the public and pr ivat e secto rs;

− (g) that procedures, processes and administrative requirements f or application, selection, par ticipation and exiting the AI

− regulator y sandbo x are sim ple, easily inte lligible, and clearly communicated in order to facilitat e the par ticipation of

− SMEs, including star t-ups, with limited leg al and administrative capacities and are streamlined across the Uni on, in order

− to av oid fragmentation and that par ticipation in an AI regulatory sandbo x established b y a Member Stat e, or b y the

− European Data Protection Super visor is mutually and unif or mly recognised and car r ies the same leg al effects across the

− Uni on;

− (h) that par ticipation in the AI regulato r y sandbo x is limited to a per iod that is appropr iat e t o the complexity and scale of

− the project and that ma y be exte nded by the national comp etent author ity ;

− (i) that AI regulator y sandbo xes f acilitate the development of to ols and infrastr ucture f or te sting, bench marking, assessing

− and explaining dimensions of AI syste ms relevant f or regulator y lear ning, such as accuracy , robustness and

− cybersecur ity , as well as measures to mitiga te r isks to fundamental r ights and society at larg e.

− 3. Prospective provider s in the AI regulatory sandbo xe s, in par ticular SMEs and star t-ups, shall be directed, where

− relevant, to pre-deplo yment ser vices such as guidance on the implementation of this Regulation, to other value-adding

− ser vices suc h as help with standardisation documents and cer tif ication, te sting and exper imentation f acilities, European

− Digital Inno vation Hubs and centres of excellence.

− 90/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− 4. Where national comp etent author ities consider author ising testing in real wo rld conditions super vised within the

− framew ork of an AI regulator y sandbo x to be established under this Ar ticle, they shall specifically agree the te r ms and

− conditions of such te sting and, in par ticular , the appropr iat e saf eguards with the par ticipants, with a view t o protecting

− fundamental r ights, health and saf ety . Where appropr iate, they shall cooperate with other national comp etent author ities

− with a view to ensur ing consistent practices across the Uni on.

− Ar ticle 59

− F ur ther processing of personal data f or dev eloping cer tain AI systems in the public interest in the AI regulator y

− sandbo x

− 1. In the AI regulato r y sandbo x, personal data lawfully collect ed f or other pur poses ma y be processed solely f or the

− pur pose of developing, training and te sting cer tain AI systems in the sandbo x when all of the f ollowing conditions are met:

− (a) AI syste ms shall be developed f or safeguarding substantial public intere st by a public author ity or another natural or

− lega l person and in one or more of the f ollo wing areas:

− (i) public safety and public health, including disease det ection, diagnosis prevention, control and treatment and

− im provement of health care systems;

− (ii) a high level of protect ion and im provement of the quality of the envir onment, protection of biodiversity , prot ection

− ag ainst pollution, green transition measures, climat e ch ange mitiga tion and adap tation measures;

− (iii) energy sustainability ;

− (iv) saf ety and resilience of transpor t systems and mobility , cr itical infrastr ucture and networks;

− (v) efficiency and quality of public administration and public ser vices;

− (b) the data processed are necessar y f or comp lying with one or more of the requirements refer red to in Chap ter III,

− Section 2 where those requirements cannot effe ctively be fulf illed b y processing anonymised , synthetic or other

− non-personal data;

− (c) there are effe ctive monitoring mechanisms to identify if any high r isks to the r ights and freedoms of the data subjects, as

− referred t o in Ar ticle 35 of Regulation (EU) 2016/679 and in Ar ticle 39 of Regulation (EU) 2018/1725, ma y ar ise

− dur ing the sandbo x exper imentation, as well as response mec hanisms to promp tly mitiga te those r isks and, where

− necessar y , stop the processing;

− (d) any personal data to be processed in the cont ext of the sandbo x are in a functionally separate , isolated and prot ected

− data processing environment under the control of the prospective pro vider and only author ised persons have access to

− those data;

− (e) provid ers can fur ther share the or iginally collect ed data only in accordance with Uni on data protection law; any

− personal data create d in the sandbo x cannot be shared outside the sandbo x;

− (f) any processing of personal data in the context of the sandbo x neither leads to measures or decisions affecting the data

− subjects nor does it affe ct the application of their r ights laid down in Uni on law on the prot ection of personal data;

− (g) any personal data processed in the cont ext of the sandbo x are protect ed by means of appropr iate te chnical and

− org anisational measures and deleted once the par ticipation in the sandbo x has te r minated or the personal data has

− reac hed the end of its ret ention per iod;

− (h) the logs of the processing of personal data in the cont ext of the sandbo x are kep t f or the duration of the par ticipation in

− the sandbo x, unless provid ed other wise by Uni on or national law;

− (i) a comp let e and detailed descr ipt ion of the process and rationale behind the training, te sting and validation of the AI

− syste m is kep t t ogether with the te sting results as par t of the t echnical documentation refer red to in Annex IV ;

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 91/144

− (j) a shor t summar y of the AI project developed in the sandbo x, its objectives and expecte d results is published on the

− website of the compet ent author ities; this obligati on shall not co ver sensitive operational data in relation to the activities

− of law enf orcement, border control, immigration or asylum author ities.

− 2. For the pur poses of the prevention, invest igation, detection or prosecution of cr iminal offe nces or the ex ecution of

− cr iminal penalties, including saf eguarding aga inst and preventing threats to public secur ity , under the control and

− responsibility of law enf orcement author ities, the processing of personal data in AI regulato r y sandbo xes shall be based on

− a specific Union or national la w and subject to the same cumulative conditions as refe r red to in paragraph 1.

− 3. Paragraph 1 is without prejudice to Uni on or national law which ex cludes processing of personal data f or other

− pur poses than those explicitly mentioned in that law , as well as to Union or national law la ying do wn the basis f or the

− processing of personal data which is necessar y f or the pur pose of developing, te sting or training of inno vative AI systems or

− any other legal basis, in comp liance with Union la w on the prot ection of personal data.

− Ar ticle 60

− T est ing of high-r isk AI sys tems in real w orld conditions outside AI regulator y sandbo xes

− 1. T esting of high-r isk AI systems in real world conditions outside AI regulator y sandbo xes ma y be conducted by

− provid ers or prospective provider s of high-r isk AI systems list ed in Annex III, in accordance with this Ar ticle and the

− real-world t esting plan referred to in this Ar ticle, without prejudice t o the prohibitions under Ar ticle 5.

− The Commission shall, b y means of implementing acts, specify the detailed elements of the real-wo rld te sting plan. Those

− imp lementing acts shall be adopt ed in accordance with the examination procedure refer red to in Ar ticle 98(2).

− This paragraph shall be without prejudice to Uni on or national law on the te sting in real w orld conditions of high-r isk AI

− syste ms related t o products covered by Union har monisation legislation list ed in Annex I.

− 2. Providers or prospective provid ers ma y conduct te sting of high-r isk AI syste ms refe r red t o in Annex III in real wo rld

− conditions at any time bef ore the placing on the mark et or the putting into ser vice of the AI syste m on their own or in

− par tnership with one or more deplo y ers or prospective deplo yers.

− 3. The te sting of high-r isk AI syste ms in real world conditions under this Ar ticle shall be without prejudice to any ethical

− review that is required by Uni on or national law .

− 4. Providers or prospective provider s ma y conduct the t esting in real world conditions only where all of the f ollowing

− conditions are met:

− (a) the provid er or prospective provider has dra wn up a real-world te sting plan and submitted it to the market sur veillance

− author ity in the Member Stat e where the te sting in real wo rld conditions is to be conducte d;

− (b) the marke t sur veillance author ity in the Member Stat e where the t esting in real world conditions is to be conducte d has

− approved the te sting in real wo rld conditions and the real-world testing plan; where the marke t sur veillance author ity

− has not provided an answer within 30 da ys, the testing in real wo rld conditions and the real-world te sting plan shall be

− understood to hav e been approved; where national law does not provid e f or a tacit approva l, the te sting in real wo rld

− conditions shall remain subject to an author isation;

− (c) the provid er or prospective provider , with the excep tion of provider s or prospective provider s of high-r isk AI syste ms

− referred to in points 1, 6 and 7 of Annex III in the areas of law enf orcement, migration, asylum and border control

− managem ent, and high-r isk AI systems referred to in point 2 of Annex III has registered the te sting in real wo rld

− conditions in accordance with Ar ticle 71(4) with a Union-wide unique sing le identification number and with the

− inf or mation specified in Annex IX; the provider or prospective provider of high-r isk AI systems refe r red to in points 1,

− 6 and 7 of Annex III in the areas of law enf orcement, migration, asylum and border control management, has register ed

− the testing in real-world conditions in the secure non-public section of the EU database according to Ar ticle 49(4), point

− (d), with a Union-wide unique sing le identifica tion number and with the inf or mation specified therein; the provid er or

− prospective provid er of high-r isk AI systems refe r red to in point 2 of Annex III has regist ered the te sting in real-wo rld

− conditions in accordance with Ar ticle 49(5);

− 92/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− (d) the provider or prospective provider conducting the testing in real wo rld conditions is established in the Union or has

− appointed a legal representative who is established in the Uni on;

− (e) data collected and processed f or the pur pose of the testing in real world conditions shall be transf er red to third

− countr ies only provided that appropr iate and applicable saf eguards under Uni on law are imp lemented;

− (f) the te sting in real w orld conditions does not last longer than necessar y t o ac hieve its objectives and in any case not

− longer than six months, which ma y be exte nded f or an additional per iod of six months, subject to pr ior notif ication by

− the pro vider or prospective provider to the marke t sur veillance author ity , accom panied b y an explanation of the need

− f or suc h an extension;

− (g) the subjects of the te sting in real wo rld conditions who are persons belonging to vulnerable groups due to their age or

− disability , are appropr iately prot ected;

− (h) where a pro vider or prospective pro vider org anises the te sting in real world conditions in cooperation with one or more

− deplo yers or prospective deplo y ers, the latter have been inf or med of all aspects of the te sting that are relevant to their

− decision to par ticipate, and given the relevant instr uctions f or use of the AI system refer red to in Ar ticle 13; the

− provid er or prospective provider and the deplo y er or prospective deplo yer shall conclude an agreement specifying their

− roles and responsibilities with a view to ensur ing comp liance with the provisions f or testing in real w orld conditions

− under this Regulation and under other applicable Uni on and national law;

− (i) the subjects of the t esting in real w orld conditions have given inf or med consent in accordance with Ar ticle 61, or in the

− case of law enf orcement, where the seeking of inf or med consent would prevent the AI system from being test ed, the

− te sting itself and the outcome of the testing in the real wo rld conditions shall not hav e an y neg ative eff ect on the

− subjects, and their personal data shall be deleted af ter the te st is perfo r med;

− (j) the te sting in real w orld conditions is effe ctively overseen by the provider or prospective pro vider , as well as by

− deplo yers or prospective deplo y ers through persons who are suitably qualified in the relevant field and have the

− necessar y capacity , training and author ity to perf or m their tasks ;

− (k) the predictions, recommendations or decisions of the AI syste m can be effectively reversed and disreg arded.

− 5. Any subjects of the te sting in real w orld conditions, or their leg ally designated representative, as appropr iate, ma y ,

− without any resulting detr iment and without having to provide any justifi cation, withdra w from the te sting at any time by

− revok ing their inf or med consent and ma y request the immediate and per manent deletion of their personal data. The

− withdra wal of the inf or med consent shall not affe ct the activities already car r ied out.

− 6. In accordance with Ar ticle 75, Member Stat es shall confer on their market sur veillance author ities the po wers of

− requir ing providers and prospective provid ers t o provide inf or mation, of car r ying out unannounced remot e or on-site

− inspections, and of perfor ming ch ecks on the conduct of the te sting in real w orld conditions and the related high-r isk AI

− syste ms. Marke t sur veillance author ities shall use those po wers to ensure the safe development of testing in real wo rld

− conditions.

− 7. Any ser ious incident identifi ed in the course of the te sting in real w orld conditions shall be repor te d to the national

− marke t sur veillance author ity in accordance with Ar ticle 73. The provid er or prospective provid er shall adopt immediate

− mitiga tion measures or , failing that, shall suspend the te sting in real world conditions until suc h mitigation take s place, or

− other wise ter minate it. The provider or prospective pro vider shall establish a procedure f or the promp t recall of the AI

− syste m upon suc h te r mination of the te sting in real world conditions.

− 8. Providers or prospective pro viders shall notify the national market sur veillance author ity in the Member Stat e where

− the te sting in real wo rld conditions is to be conducte d of the suspension or term ination of the t esting in real wo rld

− conditions and of the final outcomes.

− 9. The provid er or prospective provid er shall be liable under applicable Union and national liability law f or any damage

− caused in the course of their testing in real w orld conditions.

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 93/144

− Ar ticle 61

− Infor med consent to par ticipate in tes ting in real w orld conditions outside AI regulat or y sandbo xes

− 1. For the pur pose of te sting in real wo rld conditions under Ar ticle 60, freely-given inf or med consent shall be obtained

− from the subjects of te sting pr ior to their par ticipation in suc h te sting and af te r their hav ing been duly inf or med with

− concise, clear , relevant, and understandable inf or mation regarding:

− (a) the nature and objectives of the testing in real world conditions and the possible inconve nience that ma y be linke d to

− their par ticipation;

− (b) the conditions under which the t esting in real w orld conditions is to be conducted, including the expecte d duration of

− the subject or subjects’ par ticipation;

− (c) their r ights, and the guarantees regard ing their par ticipation, in par ticular their r ight to refuse to par ticipate in, and the

− r ight to withdraw from, testing in real wo rld conditions at any time without any resulting detr iment and without having

− to provid e any justificati on;

− (d) the ar rang ements f or requesting the reversal or the disregarding of the predictions, recommendations or decisions of

− the AI syste m;

− (e) the Uni on-wide unique sing le identification number of the t esting in real world conditions in accordance with Ar ticle

− 60(4) point (c), and the contact details of the provid er or its lega l representative from whom fur ther inf or mation can be

− obtained.

− 2. The inf or med consent shall be dated and documented and a cop y shall be given to the subjects of te sting or their leg al

− representative.

− Ar ticle 62

− Measures f or pro viders and deplo y ers, in par ticular SMEs, including st ar t-ups

− 1. Member Stat es shall under take the f ollowing actions:

− (a) provid e SMEs, including star t-ups, hav ing a registered off ice or a branc h in the Uni on, with pr ior ity access to the AI

− regulator y sandbo xe s, to the exte nt that the y fulf il the eligibility conditions and selection cr iter ia; the pr ior ity access

− shall not preclude other SMEs, including star t-ups, other than those refer red to in this paragraph from access to the AI

− regulator y sandbo x, provided that they also fulf il the eligibility conditions and selection cr ite r ia;

− (b) org anise specif ic aw areness raising and training activities on the application of this Regulation tailored to the needs of

− SMEs including star t-ups, deplo y ers and, as appropr iat e, local public author ities;

− (c) utilise existing dedicat ed channels and where appropr iate, establish new ones f or communication with SMEs including

− star t-ups, deplo yers, other innovat ors and, as appropr iate, local public author ities to provide advice and respond to

− quer ies about the implementation of this Regulation, including as rega rds par ticipation in AI regulator y sandbo xes;

− (d) f acilitate the par ticipation of SMEs and other relevant stak eholders in the standardisation development process.

− 2. The specific interests and needs of the SME provid ers, including star t-ups, shall be take n into account when setting the

− f ees f or conf or mity assessment under Ar ticle 43, reducing those f ees propor tionately to their size, marke t size and other

− relevant indicator s.

− 3. The AI Office shall under take the f ollowing actions:

− (a) provid e standardised tem plate s f or areas co vered by this Regulation, as specif ied by the Board in its request ;

− (b) develop and maintain a sing le inf or mation platf or m provid ing easy to use inf or mation in relation to this Regulation f or

− all operat ors across the Union;

− 94/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− (c) org anise appropr iate communication campaigns to raise a wareness about the obligati ons ar ising from this Regulation;

− (d) evaluate and promot e the con vergence of best practices in public procurement procedures in relation to AI systems.

− Ar ticle 63

− Derogations f or specif ic operat ors

− 1. Microenter pr ises within the meaning of Recommendation 2003/361/EC ma y comply with cer tain elements of the

− quality management system required by Ar ticle 17 of this Regulation in a simplified manner , provid ed that the y do not

− hav e par tner ent er pr ises or linked enter pr ises within the meaning of that Recommendation. For that pur pose, the

− Commission shall develop guidelines on the elements of the quality manag ement system which ma y be comp lied with in

− a simplified manner consider ing the needs of microenterp r ises, without affecting the level of protection or the need f or

− complia nce with the requirements in respect of high-r isk AI syste ms.

− 2. Paragraph 1 of this Ar ticle shall not be inter preted as ex emp ting those operat ors from fulfilling any other

− requirements or oblig ations laid do wn in this Regulation, including those established in Ar ticles 9, 10, 11, 12, 13, 14, 15,

− 72 and 73.

− C HAPTER VII

− GO VERNANCE

− SECTION 1

− Gove r nance at Union le v el

− Ar ticle 64

− AI Off ice

− 1. The Commission shall develop Union exper tise and capabilities in the field of AI through the AI Off ice.

− 2. Member Stat es shall facilitat e the tasks entr uste d t o the AI Off ice, as ref lecte d in this Regulation.

− Ar ticle 65

− Es tablishment and s tr ucture of the European Ar tif icial Intelligence Board

− 1. A European Ar tif icial Intellig ence Board (the ‘Board’) is hereb y established.

− 2. The Board shall be comp osed of one representative per Member Stat e. The European Data Protection Super visor shall

− par ticipate as obser ver . The AI Office shall also attend the Board’s meetings, without taking par t in the v otes. Other national

− and Uni on author ities, bodies or exper ts ma y be invit ed to the meetings by the Board on a case by case basis, where the

− issues discussed are of relevance f or them.

− 3. Each representative shall be designated by their Member Stat e f or a per iod of three y ears, renewable once.

− 4. Member Stat es shall ensure that their representatives on the Board:

− (a) hav e the relevant compet ences and po wers in their Member State so as to contr ibute actively to the achievement of the

− Board’s tasks refer red to in Ar ticle 66;

− (b) are designate d as a sing le contact point vis-à-vis the Board and, where appropr iate, taking into account Member States’

− needs, as a sing le contact point f or stakeholders;

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 95/144

− (c) are emp ower ed to f acilitate consistency and coordination between national compet ent author ities in their Member State

− as rega rds the imp lementation of this Regulation, including through the collection of relevant data and inf or mation f or

− the pur pose of fulf illing their tasks on the Board.

− 5. The designated representatives of the Member States shall adop t the Board’s r ules of procedure by a two-thirds

− majorit y . The r ules of procedure shall, in par ticular , la y do wn procedures f or the selection process, the duration of the

− mandate of, and specif ications of the tasks of, the Chair , detailed ar rang ements f or voting, and the organi sation of the

− Board’s activities and those of its sub-groups.

− 6. The Board shall establish two standing sub-groups to pro vide a platf or m f or cooperation and ex chang e among marke t

− sur veillance author ities and notifying author ities about issues related to market sur veillance and notified bodies respectively .

− The standing sub-group f or marke t sur veillance should act as the administrative cooperation group (ADCO) f or this

− Regulation within the meaning of Ar ticle 30 of Regulation (EU) 2019/1020.

− The Board ma y establish other standing or te mporar y sub-groups as appropr iate f or the pur pose of examining specif ic

− issues. Where appropr iate, representatives of the advisor y f or um refer red t o in Ar ticle 67 ma y be in vited to suc h sub-groups

− or t o specif ic meetings of those subgroups as obser vers.

− 7. The Board shall be org anised and operat ed so as to saf eguard the objectivity and imp ar tiality of its activities.

− 8. The Board shall be ch aired by one of the representatives of the Member Stat es. The AI Offi ce shall provid e the

− secretar iat f or the Board, con vene the meetings upon request of the Chair , and prepare the age nda in accordance with the

− tasks of the Board pursuant t o this Regulation and its r ules of procedure.

− Ar ticle 66

− T asks of the Board

− The Board shall advise and assist the Commission and the Member Stat es in order t o facilitat e the consistent and eff ective

− application of this Regulation. T o that end, the Board ma y in par ticular:

− (a) contr ibut e to the coordination among national comp etent author ities responsible f or the application of this Regulation

− and, in cooperation with and subject to the agreement of the marke t sur veillance author ities concer ned, suppor t joint

− activities of marke t sur veillance author ities refe r red to in Ar ticle 74(11);

− (b) collect and share te chnical and regulato r y exper tise and best practices among Member Stat es;

− (c) provid e advice on the imp lementation of this Regulation, in par ticular as regards the enf orcement of r ules on

− g eneral-pur pose AI models;

− (d) contr ibut e to the har monisation of administrative practices in the Member States, including in relation t o the

− derogat ion from the conf or mity assessment procedures refer red to in Ar ticle 46, the functioning of AI regulator y

− sandbo xes, and te sting in real world conditions referred to in Ar ticles 57, 59 and 60;

− (e) at the request of the Commission or on its ow n initiative, issue recommendations and wr itten opinions on an y relevant

− matt ers related to the implementation of this Regulation and to its consiste nt and effective application, including:

− (i) on the development and application of codes of conduct and codes of practice pursuant to this Regulation, as well

− as of the Commission’s guidelines;

− (ii) the evaluation and review of this Regulation pursuant to Ar ticle 112, including as rega rds the ser ious incident

− repor ts refe r red to in Ar ticle 73, and the functioning of the EU database refe r red to in Ar ticle 71, the preparation

− of the deleg ated or implementing acts, and as regard s possible alignments of this Regulation with the Uni on

− har monisation legislation list ed in Annex I;

− (iii) on tec hnical specif ications or existing standards regarding the requirements set out in Chapt er III, Section 2;

− 96/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− (iv) on the use of har monised standards or common specifications refer red to in Ar ticles 40 and 41;

− (v) trends, suc h as European g lobal comp etitiveness in AI, the uptak e of AI in the Union, and the development of

− digital skills;

− (vi) trends on the evolving typology of AI value chains, in par ticular on the resulting implications in te r ms of

− accountability ;

− (vii) on the potential need f or amendment to Annex III in accordance with Ar ticle 7, and on the potential need f or

− possible revision of Ar ticle 5 pursuant to Ar ticle 112, taking into account relevant ava ilable evidence and the

− lat est developments in te chnology ;

− (f) suppor t the Commission in promoting AI literac y , public aw areness and understanding of the benefi ts, r isks,

− saf eguards and r ights and obliga tions in relation to the use of AI syste ms;

− (g) f acilitate the development of common cr iteria and a shared understanding among market operat ors and compet ent

− author ities of the relevant concep ts provid ed f or in this Regulation, including b y contr ibuting to the development of

− bench mark s;

− (h) cooperate, as appropr iate, with other Uni on institutions, bodies, offices and ag encies, as well as relevant Uni on exper t

− groups and networks, in par ticular in the fie lds of product safety , cybersecur ity , comp etition, digital and media ser vices,

− financ ial ser vices, consumer prot ection, data and fundamental r ights prot ection;

− (i) contr ibut e to effe ctive cooperation with the compet ent author ities of third countr ies and with inter national

− org anisations;

− (j) assist national comp etent author ities and the Commission in developing the organi sational and tec hnical exper tise

− required f or the imp lementation of this Regulation, including b y contr ibuting to the assessment of training needs f or

− staff of Member States in volved in im plementing this Regulation;

− (k) assist the AI Offi ce in suppor ting national comp et ent author ities in the establishment and development of AI

− regulator y sandbo xe s, and f acilitate cooperation and inf or mation-shar ing among AI regulator y sandbo xes;

− (l) contr ibut e to, and provide relevant advice on, the development of guidance documents;

− (m) advise the Commission in relation t o inter national matt ers on AI;

− (n) provid e opinions t o the Commission on the qualif ied aler ts regard ing g eneral-pur pose AI models;

− (o) receive opinions by the Member Stat es on qualified aler ts regard ing g eneral-pur pose AI models, and on national

− exper iences and practices on the monitoring and enf orcement of AI syste ms, in par ticular systems inte grating the

− g eneral-pur pose AI models.

− Ar ticle 67

− Advisor y f or um

− 1. An advisor y f or um shall be established to provid e te chnical exper tise and advise the Board and the Commission, and

− to contr ibut e to their tasks under this Regulation.

− 2. The membership of the advisor y f or um shall represent a balanced selection of stak eholders, including industr y ,

− star t-ups, SMEs, civil society and academia. The membership of the advisor y f or um shall be balanced with rega rd to

− commercial and non-commercial interests and, within the catego r y of commercial inte rests, with regard to SMEs and other

− under takings.

− 3. The Commission shall appoint the members of the advisor y f or um, in accordance with the cr ite r ia set out in

− paragraph 2, from amongst stakeholders with recognised exper tise in the fi eld of AI.

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 97/144

− 4. The t er m of offic e of the members of the advisor y f or um shall be two years, which ma y be extended by up to no more

− than f our y ears.

− 5. The Fundamental Rights Ag ency , ENISA, the European Committe e f or Standardization (CEN), the European

− Committe e f or Electrotec hnical Standardization (CENELEC), and the European T elecommunications Standards Institute

− (ETSI) shall be per manent members of the advisor y f or um.

− 6. The advisor y f or um shall draw up its r ules of procedure. It shall elect tw o co-chairs from among its members, in

− accordance with cr iter ia set out in paragraph 2. The ter m of off ice of the co-c hairs shall be two y ears, renewable once.

− 7. The advisor y f or um shall hold meetings at least twice a y ear . The advisor y f or um ma y invit e exper ts and other

− stak eholders to its meetings.

− 8. The advisor y f or um ma y prepare opinions, recommendations and wr itt en contr ibutions at the request of the Board or

− the Commission.

− 9. The advisor y f or um ma y establish standing or tem porar y sub-groups as appropr iate f or the pur pose of examining

− specific questions related to the objectives of this Regulation.

− 10. The advisor y f or um shall prepare an annual repor t on its activities. That repor t shall be made publicly available .

− Ar ticle 68

− Scientif ic panel of independent exper ts

− 1. The Commission shall, by means of an imp lementing act, mak e provis ions on the establishment of a scientific panel

− of independent exper ts (the ‘scientific panel’) inte nded to suppor t the enf orcement activities under this Regulation. That

− imp lementing act shall be adop ted in accordance with the examination procedure refe r red to in Ar ticle 98(2).

− 2. The scientific panel shall consist of exper ts selected by the Commission on the basis of up-to -date scientific or

− te chnical exper tise in the field of AI necessar y f or the tasks set out in paragraph 3, and shall be able to demonstrate meeting

− all of the f ollowing conditions:

− (a) hav ing par ticular exper tise and compet ence and scientific or t echnical exper tise in the fie ld of AI;

− (b) independence from any provid er of AI syste ms or general-pur pose AI models;

− (c) an ability to car r y out activities dilige ntly , accurately and objectively .

− The Commission, in consultation with the Board, shall determine the number of exper ts on the panel in accordance with

− the required needs and shall ensure f air g ender and g eographical representation.

− 3. The scientifi c panel shall advise and suppor t the AI Office, in par ticular with rega rd to the f ollowing tasks:

− (a) suppor ting the im plementation and enf orcement of this Regulation as regard s g eneral-pur pose AI models and systems,

− in par ticular by:

− (i) aler ting the AI Office of possible systemic r isks at Uni on level of general-pur pose AI models, in accordance with

− Ar ticle 90;

− (ii) contr ibuting t o the development of to ols and methodologies f or evaluating capabilities of g eneral-pur pose AI

− models and syste ms, including through bench mark s;

− (iii) pro viding advice on the classif ication of g eneral-pur pose AI models with systemic r isk;

− (iv) pro viding advice on the classif ication of var ious g eneral-pur pose AI models and systems;

− 98/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− (v) contr ibuting to the development of tools and tem plates;

− (b) suppor ting the w ork of market sur veillance author ities, at their request ;

− (c) suppor ting cross-border marke t sur veillance activities as referred to in Ar ticle 74(11), without prejudice t o the po wers

− of marke t sur veillance author ities;

− (d) suppor ting the AI Offi ce in car r ying out its duties in the cont ext of the Union safeguard procedure pursuant to

− Ar ticle 81.

− 4. The exper ts on the scientific panel shall perfor m their tasks with impar tiality and objectivity , and shall ensure the

− confi dentiality of inf or mation and data obtained in car r ying out their tasks and activities. They shall neither seek nor take

− instr uctions from any one when ex ercising their tasks under paragraph 3. Each exper t shall draw up a declaration of

− intere sts, which shall be made publicly ava ilable. The AI Office shall establish systems and procedures to actively manage

− and prevent pote ntial conf licts of interest.

− 5. The im plementing act refer red to in paragraph 1 shall include provisions on the conditions, procedures and detailed

− ar rangements f or the scientific panel and its members to issue aler ts, and to request the assistance of the AI Off ice f or the

− perfo r mance of the tasks of the scientifi c panel.

− Ar ticle 69

− A ccess to the pool of exper ts by the Member States

− 1. Member States ma y call upon exper ts of the scientific panel to suppor t their enf orcement activities under this

− Regulation.

− 2. The Member States ma y be required to pa y f ees f or the advice and suppor t provid ed by the exper ts. The str ucture and

− the level of f ees as well as the scale and str ucture of reco verable costs shall be set out in the imp lementing act refer red to in

− Ar ticle 68(1), taking into account the objectives of the adequat e imp lementation of this Regulation, cost-effectiveness and

− the necessity of ensur ing effe ctive access to exper ts f or all Member States.

− 3. The Commission shall f acilitate timely access to the exper ts by the Member States, as needed, and ensure that the

− combination of suppor t activities car r ied out by Uni on AI te sting suppor t pursuant to Ar ticle 84 and exper ts pursuant to

− this Ar ticle is effi ciently organised and provides the best possible added value.

− SECTION 2

− National compe tent author ities

− Ar ticle 70

− Designation of national competent author ities and single points of cont act

− 1. Each Member State shall establish or designate as national compet ent author ities at least one notifying author ity and

− at least one marke t sur veillance author ity f or the pur poses of this Regulation. Those national comp etent author ities shall

− ex ercise their powers independently , impar tially and without bias so as to saf eguard the objectivity of their activities and

− tasks , and to ensure the application and implementation of this Regulation. The members of those author ities shall refrain

− from an y action incom patible with their duties. Provided that those pr inciples are obser ved, suc h activities and tasks ma y be

− perfo r med by one or more designate d author ities, in accordance with the organisational needs of the Member Stat e.

− 2. Member Stat es shall communicate to the Commission the identity of the notifying author ities and the marke t

− sur veillance author ities and the tasks of those author ities, as well as any subsequent ch ange s thereto . Member States shall

− make publicly availa ble inf or mation on ho w compet ent author ities and sing le points of contact can be contact ed, through

− electronic communication means by 2 Au gust 2025. Member Stat es shall designate a market sur veillance author ity to act as

− the sing le point of contact f or this Regulation, and shall notify the Commission of the identity of the sing le point of contact.

− The Commission shall make a list of the sing le points of contact publicly a vailable.

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 99/144

− 3. Member States shall ensure that their national compet ent author ities are provided with adequate tec hnical, financial

− and human resources, and with infrastr ucture to fulf il their tasks effe ctively under this Regulation. In par ticular , the national

− compet ent author ities shall hav e a suffi cient number of personnel per manently a vailable whose comp etences and exper tise

− shall include an in-depth understanding of AI te ch nologies, data and data computing , personal data prot ection,

− cybersecur ity , fundamental r ights, health and safety r isks and kno wledge of existing standards and legal requirements.

− Member States shall assess and, if necessar y , update compet ence and resource requirements refe r red to in this paragraph on

− an annual basis.

− 4. National compet ent author ities shall take appropr iate measures to ensure an adequate level of cybersecur ity .

− 5. When perfo r ming their tasks, the national com petent author ities shall act in accordance with the confi dentiality

− obliga tions set out in Ar ticle 78.

− 6. By 2 Au gust 2025, and once ever y tw o years thereaf te r , Member Stat es shall repor t to the Commission on the status

− of the financ ial and human resources of the national comp et ent author ities, with an assessment of their adequacy . The

− Commission shall transmit that inf or mation to the Board f or discussion and possible recommendations.

− 7. The Commission shall f acilitate the exc hang e of exper ience between national comp et ent author ities.

− 8. National comp etent author ities ma y provid e guidance and advice on the imp lementation of this Regulation, in

− par ticular to SMEs including star t-ups, taking into account the guidance and advice of the Board and the Commission, as

− appropr iate. Whenever national comp etent author ities intend t o provide guidance and advice with rega rd to an AI system

− in areas covered b y other Uni on law , the national comp et ent author ities under that Union la w shall be consult ed, as

− appropr iate.

− 9. Where Uni on institutions, bodies, off ices or agencies fall within the scope of this Regulation, the European Data

− Protection Super visor shall act as the compet ent author ity f or their super vision.

− CHAPTER VIII

− EU D A T ABASE FOR HIGH-RISK AI SYSTEMS

− Ar ticle 71

− EU dat abase f or high-r isk AI systems listed in Annex III

− 1. The Commission shall, in collaboration with the Member Stat es, set up and maintain an EU database containing

− inf or mation refer red to in paragraphs 2 and 3 of this Ar ticle concer ning high-r isk AI systems refe r red to in Ar ticle 6(2)

− which are registered in accordance with Ar ticles 49 and 60 and AI systems that are not considered as high-r isk pursuant to

− Ar ticle 6(3) and which are registered in accordance with Ar ticle 6(4) and Ar ticle 49. When setting the functional

− specific ations of suc h database, the Commission shall consult the relevant exper ts, and when updating the functional

− specific ations of suc h database, the Commission shall consult the Board.

− 2. The data listed in Sections A and B of Annex VIII shall be ent ered into the EU database by the provid er or , where

− applicable, by the author ised representative.

− 3. The data listed in Section C of Annex VIII shall be entered into the EU database by the deplo y er who is, or who acts on

− behalf of, a public author ity , ag ency or body , in accordance with Ar ticle 49(3) and (4).

− 4. With the ex ception of the section refe r red to in Ar ticle 49(4) and Ar ticle 60(4), point (c), the inf or mation contained in

− the EU database register ed in accordance with Ar ticle 49 shall be accessible and publicly a vailable in a user -fr iendly manner .

− The inf or mation should be easily navig able and machi ne-readable. The inf or mation regist ered in accordance with Ar ticle 60

− shall be accessible only to marke t sur veillance author ities and the Commission, unless the prospective provider or provid er

− has given consent f or also making the inf or mation accessible the public.

− 5. The EU database shall contain personal data only in so f ar as necessar y f or collecting and processing inf or mation in

− accordance with this Regulation. That inf or mation shall include the names and contact details of natural persons who are

− responsible f or registeri ng the syste m and hav e the lega l author ity t o represent the provid er or the deplo y er , as applicable.

− 100/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− 6. The Commission shall be the controller of the EU database. It shall make available to providers, prospective provider s

− and deplo y ers adequat e tec hnical and administrative suppor t. The EU database shall comply with the applicable accessibility

− requirements.

− CHAPTER IX

− POST -MARKET MONITORING, INFORMA TION SHARING AND MARKET SUR VEILL ANCE

− SECTION 1

− P ost-market monitor ing

− Ar ticle 72

− Po s t-mark et monitorin g by pro viders and pos t-mark et monitor ing plan for high-r isk AI sys tems

− 1. Providers shall establish and document a post-mark et monitoring syste m in a manner that is propor tionate to the

− nature of the AI tec hnologies and the r isks of the high-r isk AI system.

− 2. The post-mark et monito r ing system shall actively and systematically collect, document and analyse relevant data

− which ma y be provided by deplo y ers or which ma y be collected through other sources on the perform ance of high-r isk AI

− syste ms throughout their lifetime, and which allow the provider to evaluate the continuous complia nce of AI systems with

− the requirements set out in Chap ter III, Section 2. Where relevant, post-mark et monitoring shall include an analysis of the

− interac tion with other AI syste ms. This obligati on shall not co ver sensitive operational data of deplo y ers which are

− law -enf orcement author ities.

− 3. The post-mark et monitori ng system shall be based on a post-mark et monitoring plan. The post-mark et monitoring

− plan shall be par t of the tec hnical documentation referred t o in Annex IV . The Commission shall adopt an imp lementing act

− la ying do wn detailed provisions establishing a te mplat e f or the post-mark et monitoring plan and the list of elements to be

− included in the plan b y 2 Febr uar y 2026. That imp lementing act shall be adop ted in accordance with the examination

− procedure refe r red to in Ar ticle 98(2).

− 4. For high-r isk AI syste ms co vered by the Uni on har monisation legislation listed in Section A of Annex I, where

− a post-market monitoring system and plan are already established under that legislation, in order t o ensure consiste ncy ,

− av oid duplications and minimise additional burdens, provid ers shall have a choice of integrat ing, as appropr iate, the

− necessar y elements descr ibed in paragraphs 1, 2 and 3 using the te mplat e refer red in paragraph 3 into syste ms and plans

− already existing under that legislation, provid ed that it ac hieves an equivalent level of protect ion.

− The fir st subparagraph of this paragraph shall also apply to high-r isk AI syste ms refe r red to in point 5 of Annex III placed

− on the market or put into ser vice by financ ial institutions that are subject to requirements under Uni on financ ial ser vices

− law regarding their inter nal gove r nance, ar rang ements or processes.

− SECTION 2

− Shar ing of inf or mation on ser ious incidents

− Ar ticle 73

− Repor ting of ser ious incidents

− 1. Providers of high-r isk AI syste ms placed on the Union marke t shall repor t any ser ious incident to the marke t

− sur veillance author ities of the Member Stat es where that incident occur red.

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 101/144

− 2. The repor t refe r red to in paragraph 1 shall be made immediate ly af ter the provider has established a causal link

− between the AI system and the ser ious incident or the reasonable likelihood of suc h a link , and, in any event, not later than

− 15 da ys af te r the provid er or , where applicable, the deplo y er , becomes aw are of the ser ious incident.

− The per iod f or the repor ting refer red to in the fi rst subparagraph shall take account of the sever ity of the ser ious incident.

− 3. Notwithstanding paragraph 2 of this Ar ticle, in the event of a widespread infr ingem ent or a ser ious incident as

− def ined in Ar ticle 3, point (49)(b), the repor t refe r red to in paragraph 1 of this Ar ticle shall be provid ed immediately , and

− not lat er than tw o da ys af t er the provider or , where applicable, the deplo y er becomes aw are of that incident.

− 4. Notwithstanding paragraph 2, in the event of the death of a person, the repor t shall be provided immediately af te r the

− provid er or the deplo y er has established, or as soon as it suspects, a causal relationship between the high-r isk AI syste m and

− the ser ious incident, but not lat er than 10 da ys af te r the date on which the provider or , where applicable, the deplo y er

− becomes aw are of the ser ious incident.

− 5. Where necessar y to ensure timely repor ting, the pro vider or , where applicable, the deplo yer , ma y submit an initial

− repor t that is incom plete, f ollo wed by a comp let e repor t.

− 6. Following the repor ting of a ser ious incident pursuant to paragraph 1, the provid er shall, without dela y , perf or m the

− necessar y in vestigations in relation to the ser ious incident and the AI syste m concer ned. This shall include a r isk assessment

− of the incident, and cor rective action.

− The provider shall cooperate with the compet ent author ities, and where relevant with the notif ied body concer ned, dur ing

− the invest igations refer red to in the fi rst subparagraph, and shall not perfo r m any investig ation which in volves alt er ing the

− AI syste m concer ned in a wa y which ma y affe ct any subsequent evaluation of the causes of the incident, pr ior to inf or ming

− the comp et ent author ities of suc h action.

− 7. Upon receiving a notif ication related to a ser ious incident refer red to in Ar ticle 3, point (49)(c), the relevant marke t

− sur veillance author ity shall inf or m the national public author ities or bodies referred to in Ar ticle 77(1). The Commission

− shall develop dedicated guidance to f acilitate comp liance with the obligations set out in paragraph 1 of this Ar ticle. That

− guidance shall be issued by 2 August 2025, and shall be assessed regularly .

− 8. The market sur veillance author ity shall take appropr iate measures, as provid ed f or in Ar ticle 19 of Regulation (EU)

− 2019/1020, within seven da ys from the date it received the notif ication refer red to in paragraph 1 of this Ar ticle, and shall

− f ollo w the notification procedures as provided in that Regulation.

− 9. For high-r isk AI syste ms refe r red to in Annex III that are placed on the market or put into ser vice b y provider s that are

− subject to Uni on legislative instr uments la ying do wn repor ting obligati ons equivalent t o those set out in this Regulation, the

− notif ication of ser ious incidents shall be limit ed to those referred t o in Ar ticle 3, point (49)(c).

− 10. For high-r isk AI syste ms which are saf ety components of devices, or are themselves devices, co vered by Regulations

− (EU) 2017/745 and (EU) 2017/746, the notification of ser ious incidents shall be limit ed to those refer red to in Ar ticle 3,

− point (49)(c) of this Regulation, and shall be made to the national compet ent author ity chosen f or that pur pose by the

− Member Stat es where the incident occur red.

− 11. National compet ent author ities shall immediately notify the Commission of any ser ious incident, whether or not

− they hav e take n action on it, in accordance with Ar ticle 20 of Regulation (EU) 2019/1020.

− SECTION 3

− Enf or cement

− Ar ticle 74

− Marke t sur v eillance and control of AI sys tems in the Union mark et

− 1. Regulation (EU) 2019/1020 shall apply to AI systems covered b y this Regulation. For the pur poses of the eff ective

− enf orcement of this Regulation:

− 102/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− (a) any reference to an economic operator under Regulation (EU) 2019/1020 shall be understood as including all operat ors

− identifie d in Ar ticle 2(1) of this Regulation;

− (b) any reference to a product under Regulation (EU) 2019/1020 shall be understood as including all AI syste ms f alling

− within the scope of this Regulation.

− 2. As par t of their repor ting obliga tions under Ar ticle 34(4) of Regulation (EU) 2019/1020, the marke t sur veillance

− author ities shall repor t annually to the Commission and relevant national competition author ities any inf or mation

− identifie d in the course of marke t sur veillance activities that ma y be of potential interest f or the application of Uni on law on

− competition r ules. They shall also annually repor t t o the Commission about the use of prohibite d practices that occur red

− dur ing that y ear and about the measures take n.

− 3. For high-r isk AI systems related to products cover ed by the Uni on har monisation legislation listed in Section A of

− Annex I, the marke t sur veillance author ity f or the pur poses of this Regulation shall be the author ity responsible f or marke t

− sur veillance activities designated under those leg al acts.

− By derogat ion from the first subparagraph, and in appropr iate circumstances, Member Stat es ma y designate another

− relevant author ity to act as a market sur veillance author ity , provid ed the y ensure coordination with the relevant sect oral

− marke t sur veillance author ities responsible f or the enf orcement of the Union har monisation legislation listed in Annex I.

− 4. The procedures refe r red to in Ar ticles 79 to 83 of this Regulation shall not apply to AI syste ms relate d to products

− covered by the Uni on har monisation legislation listed in section A of Annex I, where such lega l acts already pro vide f or

− procedures ensur ing an equivalent level of prot ection and hav ing the same objective. In such cases, the relevant sect oral

− procedures shall apply instead.

− 5. Without prejudice to the po wers of market sur veillance author ities under Ar ticle 14 of Regulation (EU) 2019/1020,

− f or the pur pose of ensur ing the effe ctive enf orcement of this Regulation, marke t sur veillance author ities ma y ex ercise the

− powe rs refe r red to in Ar ticle 14(4), points (d) and (j), of that Regulation remotely , as appropr iat e.

− 6. For high-r isk AI syste ms placed on the marke t, put into ser vice, or used by financ ial institutions regulated by Uni on

− financ ial ser vices law , the marke t sur veillance author ity f or the pur poses of this Regulation shall be the relevant national

− author ity responsible f or the financ ial super vision of those institutions under that legislation in so f ar as the placing on the

− marke t, putting into ser vice, or the use of the AI system is in direct connection with the provision of those fi nancial

− ser vices.

− 7. By wa y of derogation from paragraph 6, in appropr iat e circumstances, and provided that coordination is ensured,

− another relevant author ity ma y be identified by the Member State as marke t sur veillance author ity f or the pur poses of this

− Regulation.

− National market sur veillance author ities super vising regulated credit institutions regulate d under Directive 2013/36/EU,

− which are par ticipating in the Sing le Super visor y Mec hanism established by Regulation (EU) No 1024/2013, should repor t,

− without dela y , to the European Central Bank an y inf or mation identified in the course of their marke t sur veillance activities

− that ma y be of potent ial interest f or the pr udential super visor y tasks of the European Central Bank specified in that

− Regulation.

− 8. For high-r isk AI syste ms listed in point 1 of Annex III to this Regulation, in so f ar as the systems are used f or la w

− enf orcement pur poses, border manag ement and justice and democracy , and f or high-r isk AI systems listed in points 6, 7

− and 8 of Annex III to this Regulation, Member States shall designate as market sur veillance author ities f or the pur poses of

− this Regulation either the comp etent data prot ection super visor y author ities under Regulation (EU) 2016/679 or Directive

− (EU) 2016/680, or any other author ity designate d pursuant to the same conditions laid down in Ar ticles 41 to 44 of

− Directive (EU) 2016/680. Market sur veillance activities shall in no wa y affect the independence of judicial author ities, or

− other wise inte r f ere with their activities when acting in their judicial capacity .

− 9. Where Uni on institutions, bodies, off ices or agencies fall within the scope of this Regulation, the European Data

− Protection Super visor shall act as their market sur veillance author ity , excep t in relation to the Cour t of Justice of the

− European Union acting in its judicial capacity .

− 10. Member States shall f acilitate coordination between marke t sur veillance author ities designate d under this Regulation

− and other relevant national author ities or bodies which super vise the application of Uni on har monisation legislation list ed

− in Annex I, or in other Uni on law , that might be relevant f or the high-r isk AI syste ms refer red to in Annex III.

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 103/144

− 11. Market sur veillance author ities and the Commission shall be able to propose joint activities, including joint

− invest igations, to be conducted by either market sur veillance author ities or mark et sur veillance author ities jointly with the

− Commission, that have the aim of promoting compliance, identifying non-compliance, raising aw areness or providing

− guidance in relation t o this Regulation with respect to specif ic cate gor ies of high-r isk AI systems that are f ound to present

− a ser ious r isk across tw o or more Member States in accordance with Ar ticle 9 of Regulation (EU) 2019/1020. The AI Off ice

− shall provid e coordination suppor t f or joint investig ations.

− 12. Without prejudice to the powers provided f or under Regulation (EU) 2019/1020, and where relevant and limited to

− what is necessar y to fulfil their tasks , the marke t sur veillance author ities shall be granted full access by provid ers to the

− documentation as well as the training, validation and te sting data sets used f or the development of high-r isk AI systems,

− including, where appropr iate and subject to secur ity saf eguards, through application programming interfaces (API) or other

− relevant te ch nical means and tools enabling remot e access.

− 13. Market sur veillance author ities shall be granted access to the source code of the high-r isk AI syste m upon a reasoned

− request and only when both of the f ollowing conditions are fulfilled:

− (a) access to source code is necessar y to assess the conf or mity of a high-r isk AI syste m with the requirements set out in

− Chapt er III, Section 2; and

− (b) te sting or auditing procedures and ver ificati ons based on the data and documentation provided by the provider have

− been exhaust ed or pro ved insuff icient.

− 14. Any inf or mation or documentation obtained by marke t sur veillance author ities shall be treate d in accordance with

− the confidentiality obligations set out in Ar ticle 78.

− Ar ticle 75

− Mutual assis tance, market sur v eillance and control of general-pur pose AI sys tems

− 1. Where an AI system is based on a general-pur pose AI model, and the model and the system are developed by the

− same provider , the AI Office shall hav e po wers t o monito r and super vise comp liance of that AI syste m with obligations

− under this Regulation. T o car r y out its monitoring and super vision tasks, the AI Offi ce shall hav e all the powers of a marke t

− sur veillance author ity pro vided f or in this Section and Regulation (EU) 2019/1020.

− 2. Where the relevant marke t sur veillance author ities hav e suffi cient reason to consider general-pur pose AI systems that

− can be used directly b y deplo y ers f or at least one pur pose that is classif ied as high-r isk pursuant t o this Regulation t o be

− non-comp liant with the requirements laid do wn in this Regulation, the y shall cooperate with the AI Office to car r y out

− complia nce evaluations, and shall inf or m the Board and other market sur veillance author ities according ly .

− 3. Where a marke t sur veillance author ity is unable to conclude its invest igation of the high-r isk AI system because of its

− inability to access cer tain inf or mation related to the general-pur pose AI model despite having made all appropr iate eff or ts

− to obtain that inf or mation, it ma y submit a reasoned request t o the AI Office, by which access to that inf or mation shall be

− enf orced. In that case, the AI Off ice shall supply to the applicant author ity without dela y , and in any event within 30 da ys,

− any inf or mation that the AI Office considers to be relevant in order to establish whether a high-r isk AI system is

− non-comp liant. Marke t sur veillance author ities shall safegua rd the confidentiality of the inf or mation that the y obtain in

− accordance with Ar ticle 78 of this Regulation. The procedure provided f or in Chapt er VI of Regulation (EU) 2019/1020

− shall apply mutatis mutandis.

− Ar ticle 76

− Super vision of testing in real w orld conditions by market sur v eillance author ities

− 1. Market sur veillance author ities shall hav e compet ences and powers to ensure that testing in real world conditions is in

− accordance with this Regulation.

− 104/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− 2. Where t esting in real wo rld conditions is conducte d f or AI systems that are super vised within an AI regulator y

− sandbo x under Ar ticle 58, the marke t sur veillance author ities shall ver ify the compliance with Ar ticle 60 as par t of their

− super visor y role f or the AI regulator y sandbo x. Those author ities ma y , as appropr iate, allo w the testing in real wo rld

− conditions to be conducte d by the provid er or prospective provid er , in deroga tion from the conditions set out in Ar ticle

− 60(4), points (f) and (g).

− 3. Where a market sur veillance author ity has been inf or med b y the prospective provider , the provid er or any third par ty

− of a ser ious incident or has other grounds f or consider ing that the conditions set out in Ar ticles 60 and 61 are not met, it

− ma y take either of the f ollowing decisions on its te r r itory , as appropr iate:

− (a) to suspend or t er minate the te sting in real world conditions;

− (b) to require the provid er or prospective provider and the deplo yer or prospective deplo y er to modify an y aspect of the

− te sting in real world conditions.

− 4. Where a marke t sur veillance author ity has take n a decision refe r red to in paragraph 3 of this Ar ticle, or has issued an

− objection within the meaning of Ar ticle 60(4), point (b), the decision or the objection shall indicate the grounds theref or

− and ho w the provid er or prospective pro vider can ch alleng e the decision or objection.

− 5. Where applicable, where a marke t sur veillance author ity has taken a decision refer red to in paragraph 3, it shall

− communicate the grounds theref or to the marke t sur veillance author ities of other Member Stat es in which the AI system

− has been test ed in accordance with the te sting plan.

− Ar ticle 77

− P o w ers of author ities protecting fundamental r ights

− 1. National public author ities or bodies which super vise or enf orce the respect of oblig ations under Uni on la w

− protect ing fundamental r ights, including the r ight to non-discr imination, in relation to the use of high-r isk AI syste ms

− referred to in Annex III shall hav e the po wer t o request and access any documentation create d or maintained under this

− Regulation in accessible language and f or mat when access t o that documentation is necessar y f or eff ectively fulf illing their

− mandates within the limits of their jur isdiction. The relevant public author ity or body shall inf or m the marke t sur veillance

− author ity of the Member State concer ned of any suc h request.

− 2. By 2 No vember 2024, each Member State shall identify the public author ities or bodies referred to in paragraph 1 and

− make a list of them publicly ava ilable. Member Stat es shall notify the list t o the Commission and to the other Member

− Stat es, and shall keep the list up to date.

− 3. Where the documentation refer red t o in paragraph 1 is insuff icient to ascer tain whether an infr ing ement of

− obliga tions under Uni on law prot ecting fundamental r ights has occur red, the public author ity or body refer red to in

− paragraph 1 ma y make a reasoned request to the marke t sur veillance author ity , to org anise te sting of the high-r isk AI

− syste m through tec hnical means. The market sur veillance author ity shall org anise the te sting with the close inv olvement of

− the requesting public author ity or body within a reasonable time f ollo wing the request.

− 4. Any inf or mation or documentation obtained by the national public author ities or bodies referred to in paragraph 1 of

− this Ar ticle pursuant to this Ar ticle shall be treated in accordance with the confidentiality obligations set out in Ar ticle 78.

− Ar ticle 78

− Conf identiality

− 1. The Commission, marke t sur veillance author ities and notif ied bodies and any other natural or lega l person in volved

− in the application of this Regulation shall, in accordance with Uni on or national law , respect the confi dentiality of

− inf or mation and data obtained in car r ying out their tasks and activities in such a manner as to prot ect, in par ticular:

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 105/144

− (a) the intellectual proper ty r ights and confidential business inf or mation or trade secrets of a natural or lega l person,

− including source code, excep t in the cases refe r red to in Ar ticle 5 of Directive (EU) 2016/943 of the European

− Parliame nt and of the Council (

− 57

− );

− (b) the eff ective imp lementation of this Regulation, in par ticular f or the pur poses of inspections, investig ations or audits;

− (c) public and national secur ity inte rests;

− (d) the conduct of cr iminal or administrative proceedings;

− (e) inf or mation classif ied pursuant to Uni on or national law .

− 2. The author ities inv olved in the application of this Regulation pursuant to paragraph 1 shall request only data that is

− str ictly necessar y f or the assessment of the r isk posed by AI syste ms and f or the exercise of their powers in accordance with

− this Regulation and with Regulation (EU) 2019/1020. They shall put in place adequate and eff ective cybersecur ity measures

− to protect the secur ity and conf identiality of the inf or mation and data obtained, and shall delete the data collected as soon

− as it is no longe r needed f or the pur pose f or which it was obtained, in accordance with applicable Union or national law .

− 3. Without prejudice to paragraphs 1 and 2, inf or mation exc hang ed on a confidential basis between the national

− compet ent author ities or between national compet ent author ities and the Commission shall not be disclosed without pr ior

− consultation of the or iginating national compet ent author ity and the deplo yer when high-r isk AI systems refe r red to in

− point 1, 6 or 7 of Annex III are used b y law enf orcement, border control, immigration or asylum author ities and when such

− disclosure wo uld jeopardise public and national secur ity interests. This ex ch ange of inf or mation shall not cover sensitive

− operational data in relation to the activities of law enf orcement, border control, immigration or asylum author ities.

− When the law enf orcement, immigration or asylum author ities are provid ers of high-r isk AI systems referred to in point 1,

− 6 or 7 of Annex III, the te chnical documentation refer red to in Annex IV shall remain within the premises of those

− author ities. Those author ities shall ensure that the mark et sur veillance author ities refe r red to in Ar ticle 74(8) and (9), as

− applicable, can, upon request, immediate ly access the documentation or obtain a cop y thereof. Only staff of the marke t

− sur veillance author ity holding the appropr iate level of secur ity clearance shall be allowed to access that documentation or

− any cop y thereof.

− 4. Paragraphs 1, 2 and 3 shall not affe ct the r ights or obliga tions of the Commission, Member States and their relevant

− author ities, as well as those of notif ied bodies, with regard to the exc hang e of inf or mation and the dissemination of

− war nings, including in the cont ext of cross-border cooperation, nor shall they affect the oblig ations of the par ties concer ned

− to provid e inf or mation under cr iminal law of the Member Stat es.

− 5. The Commission and Member Stat es ma y ex chang e, where necessar y and in accordance with relevant provisions of

− inter national and trade agreements, confi dential inf or mation with regulator y author ities of third countr ies with which they

− hav e concluded bilat eral or multilate ral conf identiality ar rang ements guarante eing an adequate level of confidentiality .

− Ar ticle 79

− Procedure at national lev el f or dealing with AI systems presenting a r isk

− 1. AI syste ms presenting a r isk shall be understood as a ‘product presenting a r isk’ as defined in Ar ticle 3, point 19 of

− Regulation (EU) 2019/1020, in so f ar as the y present r isks to the health or saf ety , or to fundamental r ights, of persons.

− 2. Where the marke t sur veillance author ity of a Member Stat e has sufficient reason to consider an AI system to present

− a r isk as refer red to in paragraph 1 of this Ar ticle, it shall car r y out an evaluation of the AI syste m concer ned in respect of

− its comp liance with all the requirements and obligations laid do wn in this Regulation. Pa r ticular attention shall be given to

− AI syste ms presenting a r isk to vulnerable groups. Where r isks to fundamental r ights are identified, the market sur veillance

− author ity shall also inf or m and fully cooperate with the relevant national public author ities or bodies referred to in Ar ticle

− 77(1). The relevant operat ors shall cooperate as necessar y with the marke t sur veillance author ity and with the other

− national public author ities or bodies refe r red to in Ar ticle 77(1).

− 106/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− (

− 57

− ) Directive (EU) 2016/943 of the European Parliament and of the Council of 8 June 2016 on the prot ection of undisclosed know-ho w

− and business inf or mation (trade secrets) against their unlaw ful acquisition, use and disclosure (OJ L 157, 15.6.2016, p. 1).

− Where, in the course of that evaluation, the market sur veillance author ity or , where applicable the market sur veillance

− author ity in cooperation with the national public author ity referred to in Ar ticle 77(1), finds that the AI system does not

− comply with the requirements and obligations laid do wn in this Regulation, it shall without undue dela y require the relevant

− operato r t o take all appropr iate cor rective actions to br ing the AI syste m into comp liance, t o withdra w the AI system from

− the mark et, or to recall it within a per iod the marke t sur veillance author ity ma y prescr ibe, and in any event within the

− shor ter of 15 wo rking da ys, or as provided f or in the relevant Uni on har monisation legislation.

− The marke t sur veillance author ity shall inf or m the relevant notif ied body according ly . Ar ticle 18 of Regulation (EU)

− 2019/1020 shall apply to the measures refe r red to in the second subparagraph of this paragraph.

− 3. Where the mark et sur veillance author ity considers that the non-comp liance is not restr ict ed to its national te r r itory , it

− shall inf or m the Commission and the other Member Stat es without undue dela y of the results of the evaluation and of the

− actions which it has required the operato r to take .

− 4. The operator shall ensure that all appropr iat e cor rective action is take n in respect of all the AI syste ms concer ned that

− it has made available on the Union marke t.

− 5. Where the operator of an AI system does not take adequate cor rective action within the per iod refer red to in

− paragraph 2, the market sur veillance author ity shall take all appropr iat e provisional measures to prohibit or restr ict the AI

− syste m’s being made available on its national marke t or put into ser vice, to withdra w the product or the standalone AI

− syste m from that marke t or to recall it. That author ity shall without undue dela y notify the Commission and the other

− Member Stat es of those measures.

− 6. The notification refer red t o in paragraph 5 shall include all ava ilable details, in par ticular the inf or mation necessar y

− f or the identification of the non-comp liant AI syste m, the or igin of the AI syste m and the supply ch ain, the nature of the

− non-comp liance alleged and the r isk in volved, the nature and duration of the national measures taken and the arguments

− put f or ward by the relevant operat or . In par ticular , the market sur veillance author ities shall indicate whether the

− non-comp liance is due to one or more of the f ollo wing:

− (a) non-comp liance with the prohibition of the AI practices refe r red t o in Ar ticle 5;

− (b) a f ailure of a high-r isk AI system to meet requirements set out in Chapt er III, Section 2;

− (c) shor tcomings in the har monised standards or common specifications refer red to in Ar ticles 40 and 41 confer r ing

− a presump tion of conf or mity ;

− (d) non-comp liance with Ar ticle 50.

− 7. The marke t sur veillance author ities other than the mark et sur veillance author ity of the Member State initiating the

− procedure shall, without undue dela y , inf or m the Commission and the other Member States of any measures adopt ed and of

− any additional inf or mation at their disposal relating to the non-compliance of the AI system concer ned, and, in the event of

− disagreement with the notif ied national measure, of their objections.

− 8. Where, within three months of receipt of the notification refe r red to in paragraph 5 of this Ar ticle, no objection has

− been raised by either a marke t sur veillance author ity of a Member State or b y the Commission in respect of a provis ional

− measure taken b y a marke t sur veillance author ity of another Member Stat e, that measure shall be deemed justifi ed. This

− shall be without prejudice to the procedural r ights of the concer ned operator in accordance with Ar ticle 18 of Regulation

− (EU) 2019/1020. The three-month per iod refe r red to in this paragraph shall be reduced to 30 da ys in the event of

− non-comp liance with the prohibition of the AI practices refe r red t o in Ar ticle 5 of this Regulation.

− 9. The marke t sur veillance author ities shall ensure that appropr iat e restr ictive measures are taken in respect of the

− product or the AI syste m concer ned, such as withdra wal of the product or the AI syste m from their marke t, without undue

− dela y .

− Ar ticle 80

− Procedure f or dealing with AI systems classif ied by the pro vider as non-high-r isk in application of Annex III

− 1. Where a mark et sur veillance author ity has sufficient reason to consider that an AI system classif ied by the provider as

− non-high-r isk pursuant to Ar ticle 6(3) is indeed high-r isk, the mark et sur veillance author ity shall car r y out an evaluation of

− the AI system concer ned in respect of its classific ation as a high-r isk AI syste m based on the conditions set out in Ar ticle

− 6(3) and the Commission guidelines.

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 107/144

− 2. Where, in the course of that evaluation, the market sur veillance author ity fi nds that the AI syste m concer ned is

− high-r isk, it shall without undue dela y require the relevant pro vider to take all necessar y actions to br ing the AI syste m into

− complia nce with the requirements and obligations laid do wn in this Regulation, as well as take appropr iate cor rective action

− within a per iod the marke t sur veillance author ity ma y prescr ibe.

− 3. Where the marke t sur veillance author ity considers that the use of the AI system concer ned is not restr icte d to its

− national terr itory , it shall inf or m the Commission and the other Member States without undue dela y of the results of the

− evaluation and of the actions which it has required the pro vider to take .

− 4. The provid er shall ensure that all necessar y action is taken to br ing the AI system into compliance with the

− requirements and obliga tions laid do wn in this Regulation. Where the provider of an AI system concer ned does not br ing

− the AI syste m into compliance with those requirements and obligati ons within the per iod refe r red to in paragraph 2 of this

− Ar ticle, the provid er shall be subject to fine s in accordance with Ar ticle 99.

− 5. The provid er shall ensure that all appropr iate cor rective action is take n in respect of all the AI systems concer ned that

− it has made available on the Union marke t.

− 6. Where the pro vider of the AI system concer ned does not take adequate cor rective action within the per iod refe r red to

− in paragraph 2 of this Ar ticle, Ar ticle 79(5) t o (9) shall apply .

− 7. Where, in the course of the evaluation pursuant to paragraph 1 of this Ar ticle, the mark et sur veillance author ity

− establishes that the AI system was misclassified by the provider as non-high-r isk in order to circum vent the application of

− requirements in Chapt er III, Section 2, the provid er shall be subject to fines in accordance with Ar ticle 99.

− 8. In ex ercising their po wer to monitor the application of this Ar ticle, and in accordance with Ar ticle 11 of Regulation

− (EU) 2019/1020, marke t sur veillance author ities ma y per f or m appropr iate ch ecks, taking into account in par ticular

− inf or mation stored in the EU database refe r red to in Ar ticle 71 of this Regulation.

− Ar ticle 81

− Union safeguard procedure

− 1. Where, within three months of receipt of the notif ication refe r red to in Ar ticle 79(5), or within 30 da ys in the case of

− non-comp liance with the prohibition of the AI practices refer red to in Ar ticle 5, objections are raised b y the marke t

− sur veillance author ity of a Member State t o a measure taken by another marke t sur veillance author ity , or where the

− Commission considers the measure to be contrar y to Uni on law , the Commission shall without undue dela y ent er into

− consultation with the marke t sur veillance author ity of the relevant Member Stat e and the operat or or operat ors, and shall

− evaluate the national measure. On the basis of the results of that evaluation, the Commission shall, within six months, or

− within 60 da ys in the case of non-comp liance with the prohibition of the AI practices refe r red to in Ar ticle 5, star ting from

− the notification refe r red t o in Ar ticle 79(5), decide whether the national measure is justified and shall notify its decision to

− the market sur veillance author ity of the Member State concer ned. The Commission shall also inf or m all other marke t

− sur veillance author ities of its decision.

− 2. Where the Commission considers the measure taken by the relevant Member Stat e to be justified, all Member Stat es

− shall ensure that they take appropr iat e restr ictive measures in respect of the AI syste m concer ned, such as requir ing the

− withdra wal of the AI system from their mark et without undue dela y , and shall inf or m the Commission according ly . Where

− the Commission considers the national measure to be unjustified, the Member Stat e concer ned shall withdra w the measure

− and shall inf or m the Commission according ly .

− 3. Where the national measure is considered justifi ed and the non-comp liance of the AI syste m is attr ibuted to

− shor tcomings in the har monised standards or common specific ations refer red to in Ar ticles 40 and 41 of this Regulation,

− the Commission shall apply the procedure provided f or in Ar ticle 11 of Regulation (EU) No 1025/2012.

− Ar ticle 82

− Compliant AI systems which present a r isk

− 1. Where, having perfor med an evaluation under Ar ticle 79, af te r consulting the relevant national public author ity

− referred to in Ar ticle 77(1), the mark et sur veillance author ity of a Member State finds that although a high-r isk AI system

− complie s with this Regulation, it never theless presents a r isk to the health or saf ety of persons, to fundamental r ights, or t o

− other aspects of public inte rest protection, it shall require the relevant operato r to take all appropr iate measures to ensure

− that the AI system concer ned, when placed on the market or put into ser vice, no longer presents that r isk without undue

− dela y , within a per iod it ma y prescr ibe.

− 108/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− 2. The provider or other relevant operato r shall ensure that cor rective action is taken in respect of all the AI syste ms

− concer ned that it has made available on the Uni on market within the timeline prescr ibed by the market sur veillance

− author ity of the Member State refer red to in paragraph 1.

− 3. The Member Stat es shall immediately inf or m the Commission and the other Member States of a finding under

− paragraph 1. That inf or mation shall include all a vailable details, in par ticular the data necessar y f or the identifica tion of the

− AI syste m concer ned, the or igin and the supply ch ain of the AI system, the nature of the r isk in v olved and the nature and

− duration of the national measures take n.

− 4. The Commission shall without undue dela y ent er into consultation with the Member Stat es concer ned and the

− relevant operators, and shall evaluate the national measures taken. On the basis of the results of that evaluation, the

− Commission shall decide whether the measure is justified and, where necessar y , propose other appropr iate measures.

− 5. The Commission shall immediately communicate its decision to the Member Stat es concer ned and to the relevant

− operato rs. It shall also inf or m the other Member States.

− Ar ticle 83

− For mal non-compliance

− 1. Where the marke t sur veillance author ity of a Member State mak es one of the f ollo wing findings, it shall require the

− relevant provid er to put an end to the non-compliance concer ned, within a per iod it ma y prescr ibe:

− (a) the CE marking has been affixed in violation of Ar ticle 48;

− (b) the CE marking has not been aff ix ed;

− (c) the EU declaration of conf or mity referred to in Ar ticle 47 has not been dra wn up;

− (d) the EU declaration of conf or mity referred to in Ar ticle 47 has not been dra wn up cor rectly ;

− (e) the registration in the EU database refe r red t o in Ar ticle 71 has not been car r ied out ;

− (f) where applicable, no author ised representative has been appointed;

− (g) te chnical documentation is not available.

− 2. Where the non-compliance refe r red to in paragraph 1 persists, the marke t sur veillance author ity of the Member State

− concer ned shall take appropr iate and propor tionate measures to restr ict or prohibit the high-r isk AI system being made

− ava ilable on the marke t or to ensure that it is recalled or withdrawn from the mark et without dela y .

− Ar ticle 84

− Union AI testing suppor t s tr uctures

− 1. The Commission shall designate one or more Union AI testing suppor t str uctures to perfo r m the tasks listed under

− Ar ticle 21(6) of Regulation (EU) 2019/1020 in the area of AI.

− 2. Without prejudice t o the tasks refer red to in paragraph 1, Uni on AI te sting suppor t str uctures shall also provide

− independent tec hnical or scientific advice at the request of the Board, the Commission, or of market sur veillance author ities.

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 109/144

− SECTION 4

− Remedies

− Ar ticle 85

− Right t o lodge a complaint with a market sur v eillance author ity

− Without prejudice t o other administrative or judicial remedies, any natural or legal person hav ing grounds to consider that

− there has been an infr ing ement of the provisions of this Regulation ma y submit complaints t o the relevant marke t

− sur veillance author ity .

− In accordance with Regulation (EU) 2019/1020, suc h complaints shall be take n into account f or the pur pose of conducting

− marke t sur veillance activities, and shall be handled in line with the dedicat ed procedures established theref or by the marke t

− sur veillance author ities.

− Ar ticle 86

− Right t o explanation of individual decision-making

− 1. Any affect ed person subject to a decision which is take n by the deplo y er on the basis of the output from a high-r isk AI

− syste m listed in Annex III, with the ex ception of syste ms listed under point 2 thereof, and which produces lega l eff ects or

− similarly signifi cantly affe cts that person in a wa y that they consider t o hav e an adverse impact on their health, saf ety or

− fundamental r ights shall hav e the r ight to obtain from the deplo y er clear and meaningful explanations of the role of the AI

− syste m in the decision-making procedure and the main elements of the decision take n.

− 2. Paragraph 1 shall not apply to the use of AI syste ms f or which excep tions from, or restr ictions to, the obliga tion

− under that paragraph f ollow from Union or national law in compliance with Union la w .

− 3. This Ar ticle shall apply only to the exte nt that the r ight refer red to in paragraph 1 is not other wise provided f or under

− Uni on law .

− Ar ticle 87

− Repor ting of infr ingements and protection of repor ting persons

− Directive (EU) 2019/1937 shall apply to the repor ting of infr ingem ents of this Regulation and the protection of persons

− repor ting suc h infr ingem ents.

− SECTION 5

− Super vision, inv estig ation, enf or cement and monitor ing in r espect of providers of g ener al-pur pose AI models

− Ar ticle 88

− Enf orcement of the obligations of pro viders of general-pur pose AI models

− 1. The Commission shall hav e ex clusive po wers t o super vise and enf orce Chapt er V , taking into account the procedural

− guarantees under Ar ticle 94. The Commission shall entr ust the imp lementation of these tasks to the AI Office, without

− prejudice to the powers of org anisation of the Commission and the division of compet ences between Member Stat es and

− the Uni on based on the T reaties.

− 2. Without prejudice to Ar ticle 75(3), marke t sur veillance author ities ma y request the Commission to ex ercise the

− powe rs laid do wn in this Section, where that is necessar y and propor tionate to assist with the fulfilment of their tasks under

− this Regulation.

− 110/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− Ar ticle 89

− Monitor ing actions

− 1. For the pur pose of car r ying out the tasks assigned to it under this Section, the AI Off ice ma y take the necessar y

− actions t o monitor the effe ctive im plementation and comp liance with this Regulation by provid ers of general-pur pose AI

− models, including their adherence to approve d codes of practice.

− 2. Downstream provid ers shall hav e the r ight t o lodg e a comp laint alleging an infr ingem ent of this Regulation.

− A comp laint shall be duly reasoned and indicate at least:

− (a) the point of contact of the provid er of the general-pur pose AI model concer ned;

− (b) a descr iption of the relevant f acts, the provisions of this Regulation concer ned, and the reason why the downstream

− provid er considers that the provid er of the general-purpo se AI model concer ned infr inged this Regulation;

− (c) any other inf or mation that the do wnstream provider that sent the request considers relevant, including, where

− appropr iate, inf or mation gathered on its ow n initiative.

− Ar ticle 90

− Aler ts of systemic r isks by the scientif ic panel

− 1. The scientifi c panel ma y provid e a qualified aler t to the AI Off ice where it has reason to suspect that:

− (a) a g eneral-pur pose AI model poses concrete identifia ble r isk at Uni on level; or

− (b) a g eneral-pur pose AI model meets the conditions refer red to in Ar ticle 51.

− 2. Upon suc h qualif ied aler t, the Commission, through the AI Offi ce and af te r hav ing inf or med the Board, ma y ex ercise

− the powers laid down in this Section f or the pur pose of assessing the matt er . The AI Office shall inf or m the Board of any

− measure according t o Ar ticles 91 t o 94.

− 3. A qualified aler t shall be duly reasoned and indicate at least:

− (a) the point of contact of the provid er of the general-pur pose AI model with systemic r isk concer ned;

− (b) a descr ipt ion of the relevant facts and the reasons f or the aler t by the scientifi c panel;

− (c) any other inf or mation that the scientific panel considers to be relevant, including, where appropr iat e, inf or mation

− g athered on its own initiative.

− Ar ticle 91

− Po w er to request document ation and inf or mation

− 1. The Commission ma y request the provider of the g eneral-pur pose AI model concer ned to provide the documentation

− dra wn up by the provider in accordance with Ar ticles 53 and 55, or any additional inf or mation that is necessar y f or the

− pur pose of assessing comp liance of the pro vider with this Regulation.

− 2. Bef ore sending the request f or inf or mation, the AI Offi ce ma y initiate a str uctured dialogue with the provid er of the

− g eneral-pur pose AI model.

− 3. Upon a duly substantiated request from the scientific panel, the Commission ma y issue a request f or inf or mation to

− a provider of a g eneral-pur pose AI model, where the access to inf or mation is necessar y and propor tionate f or the fulfilment

− of the tasks of the scientific panel under Ar ticle 68(2).

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 111/144

− 4. The request f or inf or mation shall state the lega l basis and the pur pose of the request, specify what inf or mation is

− required, set a per iod within which the inf or mation is to be provid ed, and indicate the fine s provided f or in Ar ticle 101 f or

− supplying incor rect, incom plete or misleading inf or mation.

− 5. The provider of the g eneral-pur pose AI model concer ned, or its representative shall supply the inf or mation requested.

− In the case of lega l persons, comp anies or fir ms, or where the provider has no lega l personality , the persons author ised to

− represent them by la w or by their statutes, shall supply the inf or mation requested on behalf of the provider of the

− g eneral-pur pose AI model concer ned. Lawy ers duly author ised to act ma y supply inf or mation on behalf of their clients. The

− clients shall never theless remain fully responsible if the inf or mation supplied is incomplet e, incor rect or misleading.

− Ar ticle 92

− Po w er to conduct evaluations

− 1. The AI Office, af ter consulting the Board, ma y conduct evaluations of the g eneral-pur pose AI model concer ned:

− (a) to assess comp liance of the provid er with obliga tions under this Regulation, where the inf or mation gathered pursuant

− to Ar ticle 91 is insufficient ; or

− (b) to investig ate syste mic r isk s at Uni on level of g eneral-pur pose AI models with syste mic r isk, in par ticular f ollowing

− a qualified aler t from the scientific panel in accordance with Ar ticle 90(1), point (a).

− 2. The Commission ma y decide to appoint independent exper ts to car r y out evaluations on its behalf, including from the

− scientific panel established pursuant to Ar ticle 68. Independent exper ts appointed f or this task shall meet the cr ite r ia

− outlined in Ar ticle 68(2).

− 3. For the pur poses of paragraph 1, the Commission ma y request access to the g eneral-pur pose AI model concer ned

− through APIs or fur ther appropr iat e te chnical means and to ols, including source code.

− 4. The request f or access shall state the lega l basis, the pur pose and reasons of the request and set the per iod within

− which the access is to be provided, and the fines provided f or in Ar ticle 101 f or f ailure to provide access.

− 5. The pro viders of the general-purpo se AI model concer ned or its representative shall supply the inf or mation requested.

− In the case of lega l persons, comp anies or fir ms, or where the provider has no lega l personality , the persons author ised to

− represent them by law or b y their statute s, shall provide the access request ed on behalf of the pro vider of the

− g eneral-pur pose AI model concer ned.

− 6. The Commission shall adopt imp lementing acts setting out the detailed ar rang ements and the conditions f or the

− evaluations, including the detailed ar rang ements f or inv olving independent exper ts, and the procedure f or the selection

− thereof. Those im plementing acts shall be adop ted in accordance with the examination procedure referred to in Ar ticle

− 98(2).

− 7. Pr ior to requesting access t o the g eneral-pur pose AI model concer ned, the AI Offi ce ma y initiate a str uctured dialogue

− with the pro vider of the g eneral-pur pose AI model to gather more inf or mation on the inter nal t esting of the model, intern al

− saf eguards f or preventing systemic r isks, and other inte r nal procedures and measures the provid er has taken to mitigat e

− suc h r isks.

− Ar ticle 93

− Po w er to request measures

− 1. Where necessar y and appropr iat e, the Commission ma y request provid ers to :

− (a) take appropr iate measures to comp ly with the obligations set out in Ar ticles 53 and 54;

− 112/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− (b) imp lement mitig ation measures, where the evaluation car r ied out in accordance with Ar ticle 92 has given r ise to ser ious

− and substantiate d concer n of a syste mic r isk at Union level;

− (c) restr ict the making available on the market, withdra w or recall the model.

− 2. Bef ore a measure is requested, the AI Office ma y initiate a str uctured dialogue with the provider of the

− g eneral-pur pose AI model.

− 3. If, dur ing the str uctured dialogue refe r red t o in paragraph 2, the pro vider of the gene ral-pur pose AI model with

− syste mic r isk offers commitments to implement mitigation measures to address a syste mic r isk at Union level, the

− Commission ma y , b y decision, make those commitments binding and declare that there are no fur ther grounds f or action.

− Ar ticle 94

− Procedural r ights of economic operat ors of the general-pur pose AI model

− Ar ticle 18 of Regulation (EU) 2019/1020 shall apply mutatis mutandis t o the providers of the general-pur pose AI model,

− without prejudice to more specific procedural r ights provided f or in this Regulation.

− C HAPTER X

− CODES OF CONDUCT AND GUIDELINES

− Ar ticle 95

− Codes of conduct for v olunt ar y application of specif ic requirements

− 1. The AI Offi ce and the Member States shall encourage and facilitat e the drawing up of codes of conduct, including

− related gover nance mechanisms, inte nded to f oster the v oluntar y application to AI systems, other than high-r isk AI systems,

− of some or all of the requirements set out in Chap ter III, Section 2 taking into account the a vailable tec hnical solutions and

− industr y best practices allowi ng f or the application of suc h requirements.

− 2. The AI Office and the Member States shall f acilitate the dra wing up of codes of conduct concer ning the v oluntar y

− application, including by deplo y ers, of specific requirements t o all AI syste ms, on the basis of clear objectives and k ey

− perfo r mance indicato rs to measure the achievement of those objectives, including elements suc h as, but not limited t o:

− (a) applicable elements provided f or in Uni on ethical guidelines f or tr ustwo r th y AI;

− (b) assessing and minimising the imp act of AI syste ms on envir onmental sustainability , including as regards energy-eff icient

− programming and te chniques f or the efficient design, training and use of AI;

− (c) promoting AI literac y , in par ticular that of persons dealing with the development, operation and use of AI;

− (d) f acilitating an inclusive and diverse design of AI syste ms, including through the establishment of inclusive and diverse

− development te ams and the promotion of stak eholders’ par ticipation in that process;

− (e) assessing and preventing the nega tive imp act of AI syste ms on vulnerable persons or groups of vulnerable persons,

− including as regards accessibility f or persons with a disability , as well as on g ender equality .

− 3. Codes of conduct ma y be drawn up by individual provid ers or deplo y ers of AI syste ms or b y organisations

− representing them or b y both, including with the inv olvement of any intere sted stakeholders and their representative

− org anisations, including civil society org anisations and academia. Codes of conduct ma y cover one or more AI syste ms

− taking into account the similar ity of the intende d pur pose of the relevant systems.

− 4. The AI Office and the Member Stat es shall take into account the specific interests and needs of SMEs, including

− star t-ups, when encouraging and f acilitating the dra wing up of codes of conduct.

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 113/144

− Ar ticle 96

− Guidelines from the Commission on the implementa tion of this Regulation

− 1. The Commission shall develop guidelines on the practical imp lementation of this Regulation, and in par ticular on:

− (a) the application of the requirements and obliga tions refe r red to in Ar ticles 8 to 15 and in Ar ticle 25;

− (b) the prohibite d practices refe r red to in Ar ticle 5;

− (c) the practical implementation of the provisions relate d to substantial modification;

− (d) the practical implementation of transparency obliga tions laid do wn in Ar ticle 50;

− (e) detailed inf or mation on the relationship of this Regulation with the Union har monisation legislation listed in Annex I,

− as well as with other relevant Uni on law , including as regards consistency in their enf orcement ;

− (f) the application of the definition of an AI system as set out in Ar ticle 3, point (1).

− When issuing suc h guidelines, the Commission shall pa y par ticular atte ntion to the needs of SMEs including star t-ups, of

− local public author ities and of the sect ors most likely to be affe cted by this Regulation.

− The guidelines refe r red to in the first subparagraph of this paragraph shall take due account of the g enerally ac knowledg ed

− state of the ar t on AI, as well as of relevant har monised standards and common specif ications that are refer red to in

− Ar ticles 40 and 41, or of those har monised standards or te chnical specific ations that are set out pursuant to Uni on

− har monisation law .

− 2. At the request of the Member States or the AI Office, or on its own initiative, the Commission shall update guidelines

− previously adop ted when deemed necessar y .

− CHAPTER XI

− DELEGA TION OF PO WER AND COMM ITTEE PR OCEDURE

− Ar ticle 97

− Exe rcise of the delegation

− 1. The po wer to adop t deleg ated acts is confe r red on the Commission subject to the conditions laid do wn in this Ar ticle.

− 2. The po wer to adop t deleg ated acts refe r red to in Ar ticle 6(6) and (7), Ar ticle 7(1) and (3), Ar ticle 11(3), Ar ticle 43(5)

− and (6), Ar ticle 47(5), Ar ticle 51(3), Ar ticle 52(4) and Ar ticle 53(5) and (6) shall be conferr ed on the Commission f or

− a per iod of five y ears from 1 August 2024. The Commission shall dra w up a repor t in respect of the delegation of power

− not lat er than nine months bef ore the end of the five-year per iod. The deleg ation of power shall be tacitly extende d f or

− per iods of an identical duration, unless the European Parliame nt or the Council opposes suc h extension not later than three

− months bef ore the end of each per iod.

− 3. The delegation of power refe r red to in Ar ticle 6(6) and (7), Ar ticle 7(1) and (3), Ar ticle 11(3), Ar ticle 43(5) and (6),

− Ar ticle 47(5), Ar ticle 51(3), Ar ticle 52(4) and Ar ticle 53(5) and (6) ma y be revok ed at an y time by the European Pa rliament

− or b y the Council. A decision of revoc ation shall put an end to the deleg ation of power specif ied in that decision. It shall

− take effect the da y f ollowing that of its publication in the Off icial Jour nal of the Eur opean Union or at a later date specified

− therein. It shall not affe ct the validity of any delegat ed acts already in f orce.

− 4. Bef ore adopting a delegat ed act, the Commission shall consult exper ts designate d by each Member Stat e in accordance

− with the pr inciples laid do wn in the Inte r institutional Agreement of 13 Apr il 2016 on Bett er Law-Making .

− 114/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− 5. As soon as it adop ts a deleg ated act, the Commission shall notify it simultaneously t o the European Pa rliament and to

− the Council.

− 6. Any delegat ed act adop te d pursuant to Ar ticle 6(6) or (7), Ar ticle 7(1) or (3), Ar ticle 11(3), Ar ticle 43(5) or (6),

− Ar ticle 47(5), Ar ticle 51(3), Ar ticle 52(4) or Ar ticle 53(5) or (6) shall ent er into f orce only if no objection has been

− expressed by either the European Parliame nt or the Council within a per iod of three months of notification of that act to

− the European Pa rliament and the Council or if, bef ore the expir y of that per iod, the European Parliament and the Council

− hav e both inf or med the Commission that they will not object. That per iod shall be extended by three months at the

− initiative of the European Parliame nt or of the Council.

− Ar ticle 98

− Committee procedure

− 1. The Commission shall be assist ed by a committee. That committee shall be a committ ee within the meaning of

− Regulation (EU) No 182/2011.

− 2. Where refe rence is made to this paragraph, Ar ticle 5 of Regulation (EU) No 182/2011 shall apply .

− C HAPTER XII

− PENAL TIES

− Ar ticle 99

− Pe nalties

− 1. In accordance with the te r ms and conditions laid down in this Regulation, Member Stat es shall la y do wn the r ules on

− penalties and other enf orcement measures, which ma y also include war nings and non-monetar y measures, applicable to

− infr ingements of this Regulation b y operat ors, and shall take all measures necessar y to ensure that the y are properly and

− effe ctively implement ed, thereby taking into account the guidelines issued b y the Commission pursuant to Ar ticle 96. The

− penalties provided f or shall be effe ctive, propor tionate and dissuasive. They shall tak e into account the interests of SMEs,

− including star t-ups, and their economic viability .

− 2. The Member States shall, without dela y and at the lat est by the date of entr y into application, notify the Commission

− of the r ules on penalties and of other enf orcement measures refe r red to in paragraph 1, and shall notify it, without dela y , of

− any subsequent amendment to them.

− 3. Non-compliance with the prohibition of the AI practices refe r red to in Ar ticle 5 shall be subject t o administrative

− fine s of up t o EUR 35 000 000 or , if the offe nder is an under taking, up t o 7 % of its to tal wo rldwide annual tur nover f or the

− preceding fi nancial y ear , whichever is higher .

− 4. Non-compliance with any of the f ollowi ng provisions relate d to operato rs or notified bodies, other than those laid

− do wn in Ar ticles 5, shall be subject to administrative fines of up to EUR 15 000 000 or , if the offe nder is an under taking, up

− to 3 % of its total worldwide annual tur nover f or the preceding financ ial y ear , whichever is higher:

− (a) obliga tions of provid ers pursuant to Ar ticle 16;

− (b) obliga tions of author ised representatives pursuant t o Ar ticle 22;

− (c) obliga tions of im por ters pursuant to Ar ticle 23;

− (d) obliga tions of distr ibutors pursuant to Ar ticle 24;

− (e) obliga tions of deplo yers pursuant to Ar ticle 26;

− (f) requirements and obliga tions of notified bodies pursuant to Ar ticle 31, Ar ticle 33(1), (3) and (4) or Ar ticle 34;

− (g) transparency obliga tions f or provid ers and deplo yers pursuant to Ar ticle 50.

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 115/144

− 5. The supply of incor rect, incom plete or misleading inf or mation to notif ied bodies or national compet ent author ities in

− reply to a request shall be subject to administrative fi nes of up to EUR 7 500 000 or , if the offe nder is an under taking, up t o

− 1 % of its total worldwide annual tur nover f or the preceding fi nancial y ear , whichever is higher .

− 6. In the case of SMEs, including star t-ups, each fine referred to in this Ar ticle shall be up to the percentage s or amount

− referred to in paragraphs 3, 4 and 5, whichever thereof is low er .

− 7. When deciding whether to im pose an administrative fine and when deciding on the amount of the administrative fi ne

− in each individual case, all relevant circumstances of the specif ic situation shall be take n into account and, as appropr iate,

− regard shall be given to the f ollowing:

− (a) the nature, gra vity and duration of the infr ingem ent and of its consequences, taking into account the pur pose of the AI

− syste m, as well as, where appropr iate, the number of affect ed persons and the level of damage suffer ed by them;

− (b) whether administrative fines have already been applied by other market sur veillance author ities to the same operator f or

− the same infr ingem ent ;

− (c) whether administrative fine s hav e already been applied by other author ities to the same operato r f or infr ing ements of

− other Uni on or national law , when suc h infr ing ements result from the same activity or omission constituting a relevant

− infr ingement of this Regulation;

− (d) the size, the annual tur nover and marke t share of the operat or committing the infr ing ement ;

− (e) any other aggravating or mitiga ting fact or applicable t o the circumstances of the case, suc h as financ ial benefits gained,

− or losses av oided, directly or indirectly , from the infr ingem ent ;

− (f) the degree of cooperation with the national compet ent author ities, in order to remedy the infr ingem ent and mitigat e the

− possible adverse effe cts of the infr ingement;

− (g) the degree of responsibility of the operator taking into account the te chnical and org anisational measures imp lemented

− by it ;

− (h) the manner in which the infr ing ement became kno wn t o the national comp etent author ities, in par ticular whether , and

− if so to what exte nt, the operato r notif ied the infr ingem ent ;

− (i) the inte ntional or neg lig ent ch aracter of the infr ingement;

− (j) any action taken by the operator t o mitigat e the har m suffer ed by the affe cted persons.

− 8. Each Member State shall la y do wn r ules on to what extent administrative fi nes ma y be imp osed on public author ities

− and bodies established in that Member State.

− 9. Depending on the leg al syste m of the Member States, the r ules on administrative fine s ma y be applied in such

− a manner that the fines are imp osed by compet ent national cour ts or by other bodies, as applicable in those Member Stat es.

− The application of suc h r ules in those Member Stat es shall hav e an equivalent effect.

− 10. The ex ercise of powers under this Ar ticle shall be subject to appropr iate procedural saf eguards in accordance with

− Uni on and national law , including eff ective judicial remedies and due process.

− 11. Member Stat es shall, on an annual basis, repor t to the Commission about the administrative fines the y have issued

− dur ing that y ear , in accordance with this Ar ticle, and about any related litigation or judicial proceedings.

− Ar ticle 100

− A dminis trativ e f ines on Union instit utions, bodies, off ices and agencies

− 1. The European Data Protection Super visor ma y imp ose administrative fine s on Uni on institutions, bodies, offices and

− age ncies f alling within the scope of this Regulation. When deciding whether to im pose an administrative fine and when

− deciding on the amount of the administrative fine in each individual case, all relevant circumstances of the specific situation

− shall be take n into account and due rega rd shall be given to the f ollowi ng:

− 116/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− (a) the nature, gravity and duration of the infr ingem ent and of its consequences, taking into account the pur pose of the AI

− syste m concer ned, as well as, where appropr iate, the number of affe cted persons and the level of damage suff ered by

− them;

− (b) the degree of responsibility of the Uni on institution, body , off ice or ag ency , taking into account te ch nical and

− org anisational measures imp lemented by them;

− (c) any action take n b y the Uni on institution, body , office or age ncy to mitigat e the damage suffered by affe cted persons;

− (d) the degree of cooperation with the European Data Prot ection Super visor in order to remedy the infr ing ement and

− mitiga te the possible adverse effe cts of the infr ing ement, including comp liance with any of the measures previously

− ordered b y the European Data Prot ection Super visor against the Union institution, body , off ice or age ncy concer ned

− with regard t o the same subject matt er ;

− (e) any similar previous infr ingements by the Uni on institution, body , off ice or age ncy ;

− (f) the manner in which the infr ingem ent became kno wn to the European Data Prot ection Super visor , in par ticular

− whether , and if so to what extent, the Uni on institution, body , off ice or age ncy notif ied the infr ingem ent ;

− (g) the annual budg et of the Uni on institution, body , off ice or age ncy .

− 2. Non-compliance with the prohibition of the AI practices refe r red to in Ar ticle 5 shall be subject t o administrative

− fine s of up to EUR 1 500 000.

− 3. The non-com pliance of the AI syste m with any requirements or obligati ons under this Regulation, other than those

− laid down in Ar ticle 5, shall be subject to administrative fine s of up to EUR 750 000.

− 4. Bef ore taking decisions pursuant to this Ar ticle, the European Data Protect ion Super visor shall give the Uni on

− institution, body , off ice or agency which is the subject of the proceedings conducted b y the European Data Protection

− Super visor the oppor tunity of being heard on the matt er regard ing the possible infr ingem ent. The European Data

− Protection Super visor shall base his or her decisions only on elements and circumstances on which the par ties concer ned

− hav e been able to comment. Comp lainants, if an y , shall be associated closely with the proceedings.

− 5. The r ights of defence of the par ties concer ned shall be fully respected in the proceedings. They shall be entitled to

− hav e access to the European Data Protection Super visor ’s fi le, subject to the legitimat e interest of individuals or

− under takings in the protect ion of their personal data or business secrets.

− 6. Funds collected by imp osition of fine s in this Ar ticle shall contr ibut e to the g eneral budg et of the Uni on. The fine s

− shall not affect the effe ctive operation of the Uni on institution, body , off ice or age ncy fine d.

− 7. The European Data Prot ection Super visor shall, on an annual basis, notify the Commission of the administrative fine s

− it has imp osed pursuant t o this Ar ticle and of any litig ation or judicial proceedings it has initiated.

− Ar ticle 101

− Fin es for pro viders of general-pur pose AI models

− 1. The Commission ma y impose on provid ers of general-pur pose AI models fines not ex ceeding 3 % of their annual total

− wo rldwide tur nove r in the preceding financial y ear or EUR 15 000 000, whichever is higher ., when the Commission finds

− that the provider intent ionally or neg ligently:

− (a) infr inged the relevant provisions of this Regulation;

− (b) f ailed to comply with a request f or a document or f or inf or mation pursuant to Ar ticle 91, or supplied incor rect,

− incom plete or misleading inf or mation;

− (c) f ailed to comp ly with a measure request ed under Ar ticle 93;

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 117/144

− (d) f ailed to make available to the Commission access to the g eneral-pur pose AI model or g eneral-pur pose AI model with

− syste mic r isk with a view to conducting an evaluation pursuant to Ar ticle 92.

− In fixing the amount of the fine or per iodic penalty pa yment, regard shall be had to the nature, gra vity and duration of the

− infr ingement, taking due account of the pr inciples of propor tionality and appropr iat eness. The Commission shall also into

− account commitments made in accordance with Ar ticle 93(3) or made in relevant codes of practice in accordance with

− Ar ticle 56.

− 2. Bef ore adop ting the decision pursuant to paragraph 1, the Commission shall communicate its preliminar y findings to

− the provid er of the gene ral-pur pose AI model and give it an oppor tunity to be heard.

− 3. Fines im posed in accordance with this Ar ticle shall be effe ctive, propor tionate and dissuasive.

− 4. Inf or mation on fi nes imp osed under this Ar ticle shall also be communicated to the Board as appropr iate.

− 5. The Cour t of Justice of the European Uni on shall hav e unlimit ed jur isdiction to review decisions of the Commission

− fixing a fi ne under this Ar ticle. It ma y cancel, reduce or increase the fi ne imposed.

− 6. The Commission shall adop t imp lementing acts containing detailed ar rang ements and procedural saf eguards f or

− proceedings in view of the possible adop tion of decisions pursuant to paragraph 1 of this Ar ticle. Those imp lementing acts

− shall be adop ted in accordance with the examination procedure refer red to in Ar ticle 98(2).

− C HAPTER XIII

− FINAL PR O VISIONS

− Ar ticle 102

− Amendment to Regulation (EC) No 300/2008

− In Ar ticle 4(3) of Regulation (EC) No 300/2008, the f ollowing subparagraph is added:

− ‘ When adopti ng detailed measures related to te chnical specif ications and procedures f or approva l and use of secur ity

− equipment concer ning Ar tificial Intellig ence syste ms within the meaning of Regulation (EU) 2024/1689 of the European

− Parliame nt and of the Council (*), the requirements set out in Chap ter III, Section 2, of that Regulation shall be taken into

− account.

− (*) Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 la ying down har monised

− r ules on ar tif icial intellig ence and amending Regulations (EC) No 300/2008, (EU) No 167/2013, (EU) No 168/2013,

− (EU) 2018/858, (EU) 2018/1139 and (EU) 2019/2144 and Directives 2014/90/EU, (EU) 2016/797 and (EU)

− 2020/1828 (Ar tificial Inte lligence A ct) (OJ L, 2024/1689, 12.7.2024, ELI: http://data.europa.eu/eli/reg/

− 2024/1689/oj).’ .

− Ar ticle 103

− Amendment to Regulation (EU) No 167/2013

− In Ar ticle 17(5) of Regulation (EU) No 167/2013, the f ollo wing subparagraph is added:

− ‘ When adopting delegat ed acts pursuant to the first subparagraph concer ning ar tif icial intellig ence syste ms which are safety

− compo nents within the meaning of Regulation (EU) 2024/1689 of the European Pa rliament and of the Council (*), the

− requirements set out in Chap ter III, Section 2, of that Regulation shall be taken into account.

− (*) Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 la ying down har monised

− r ules on ar tif icial intellig ence and amending Regulations (EC) No 300/2008, (EU) No 167/2013, (EU) No 168/2013,

− (EU) 2018/858, (EU) 2018/1139 and (EU) 2019/2144 and Directives 2014/90/EU, (EU) 2016/797 and (EU)

− 2020/1828 (Ar tificial Inte lligence A ct) (OJ L, 2024/1689, 12.7.2024, ELI: http://data.europa.eu/eli/reg/

− 2024/1689/oj).’ .

− 118/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− Ar ticle 104

− Amendment to Regulation (EU) No 168/2013

− In Ar ticle 22(5) of Regulation (EU) No 168/2013, the f ollo wing subparagraph is added:

− ‘ When adopting deleg ated acts pursuant to the first subparagraph concer ning Ar tif icial Intell igence syste ms which are saf ety

− compo nents within the meaning of Regulation (EU) 2024/1689 of the European Pa rliament and of the Council (*), the

− requirements set out in Chap ter III, Section 2, of that Regulation shall be taken into account.

− (*) Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 la ying down har monised

− r ules on ar tif icial intellig ence and amending Regulations (EC) No 300/2008, (EU) No 167/2013, (EU) No 168/2013,

− (EU) 2018/858, (EU) 2018/1139 and (EU) 2019/2144 and Directives 2014/90/EU, (EU) 2016/797 and (EU)

− 2020/1828 (Ar tificial Inte lligence A ct) (OJ L, 2024/1689, 12.7.2024, ELI: http://data.europa.eu/eli/reg/

− 2024/1689/oj).’ .

− Ar ticle 105

− Amendment to Directiv e 2014/90/EU

− In Ar ticle 8 of Directive 2014/90/EU, the f ollowing paragraph is added:

− ‘5. For Ar tif icial Inte lligence syste ms which are safety comp onents within the meaning of Regulation (EU) 2024/1689 of

− the European Pa rliament and of the Council (*), when car r ying out its activities pursuant to paragraph 1 and when adop ting

− te chnical specif ications and t esting standards in accordance with paragraphs 2 and 3, the Commission shall take into

− account the requirements set out in Chap ter III, Section 2, of that Regulation.

− (*) Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 la ying down har monised

− r ules on ar tif icial intellig ence and amending Regulations (EC) No 300/2008, (EU) No 167/2013, (EU) No 168/2013,

− (EU) 2018/858, (EU) 2018/1139 and (EU) 2019/2144 and Directives 2014/90/EU, (EU) 2016/797 and (EU)

− 2020/1828 (Ar tificial Inte lligence A ct) (OJ L, 2024/1689, 12.7.2024, ELI: http://data.europa.eu/eli/reg/

− 2024/1689/oj).’ .

− Ar ticle 106

− Amendment to Directiv e (EU) 2016/797

− In Ar ticle 5 of Directive (EU) 2016/797, the f ollo wing paragraph is added:

− ‘12. When adop ting delegat ed acts pursuant to paragraph 1 and im plementing acts pursuant to paragraph 11

− concer ning Ar tificial Intellig ence systems which are safety components within the meaning of Regulation (EU) 2024/1689

− of the European Pa rliament and of the Council (*), the requirements set out in Chapt er III, Section 2, of that Regulation shall

− be take n into account.

− (*) Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 la ying down har monised

− r ules on ar tif icial intellig ence and amending Regulations (EC) No 300/2008, (EU) No 167/2013, (EU) No 168/2013,

− (EU) 2018/858, (EU) 2018/1139 and (EU) 2019/2144 and Directives 2014/90/EU, (EU) 2016/797 and (EU)

− 2020/1828 (Ar tificial Inte lligence A ct) (OJ L, 2024/1689, 12.7.2024, ELI: http://data.europa.eu/eli/reg/

− 2024/1689/oj).’ .

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 119/144

− Ar ticle 107

− Amendment to Regulation (EU) 2018/858

− In Ar ticle 5 of Regulation (EU) 2018/858 the f ollowi ng paragraph is added:

− ‘4. When adop ting deleg ated acts pursuant to paragraph 3 concer ning Ar tif icial Intellig ence syste ms which are safety

− compo nents within the meaning of Regulation (EU) 2024/1689 of the European Pa rliament and of the Council (*), the

− requirements set out in Chap ter III, Section 2, of that Regulation shall be taken into account.

− (*) Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 la ying down har monised

− r ules on ar tif icial intellig ence and amending Regulations (EC) No 300/2008, (EU) No 167/2013, (EU) No 168/2013,

− (EU) 2018/858, (EU) 2018/1139 and (EU) 2019/2144 and Directives 2014/90/EU, (EU) 2016/797 and (EU)

− 2020/1828 (Ar tificial Inte lligence A ct) (OJ L, 2024/1689, 12.7.2024, ELI: http://data.europa.eu/eli/reg/

− 2024/1689/oj).’ .

− Ar ticle 108

− Amendments t o Regulation (EU) 2018/1139

− Regulation (EU) 2018/1139 is amended as f ollo ws:

− (1) in Ar ticle 17, the f ollowing paragraph is added:

− ‘3. Without prejudice to paragraph 2, when adop ting im plementing acts pursuant to paragraph 1 concer ning

− Ar tif icial Intellig ence systems which are safety comp onents within the meaning of Regulation (EU) 2024/1689 of the

− European Parliame nt and of the Council (*), the requirements set out in Chapt er III, Section 2, of that Regulation shall be

− tak en into account.

− (*) Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 la ying do wn

− har monised r ules on ar tif icial intellig ence and amending Regulations (EC) No 300/2008, (EU) No 167/2013, (EU)

− No 168/2013, (EU) 2018/858, (EU) 2018/1139 and (EU) 2019/2144 and Directives 2014/90/EU, (EU) 2016/797

− and (EU) 2020/1828 (Ar tif icial Intellig ence A ct) (OJ L, 2024/1689, 12.7.2024, ELI: http://data.europa.

− eu/eli/reg/2024/1689/oj).’;

− (2) in Ar ticle 19, the f ollowing paragraph is added:

− ‘4. When adop ting deleg ated acts pursuant to paragraphs 1 and 2 concer ning Ar tif icial Intel ligence syste ms which

− are saf ety comp onents within the meaning of Regulation (EU) 2024/1689, the requirements set out in Chapt er III,

− Section 2, of that Regulation shall be taken into account.’;

− (3) in Ar ticle 43, the f ollowing paragraph is added:

− ‘4. When adop ting im plementing acts pursuant to paragraph 1 concer ning Ar tif icial Inte lligence systems which are

− saf ety compo nents within the meaning of Regulation (EU) 2024/1689, the requirements set out in Chapt er III,

− Section 2, of that Regulation shall be taken into account.’;

− (4) in Ar ticle 47, the f ollowing paragraph is added:

− ‘3. When adop ting deleg ated acts pursuant to paragraphs 1 and 2 concer ning Ar tif icial Intel ligence syste ms which

− are saf ety comp onents within the meaning of Regulation (EU) 2024/1689, the requirements set out in Chapt er III,

− Section 2, of that Regulation shall be taken into account.’;

− (5) in Ar ticle 57, the f ollowing subparagraph is added:

− ‘ When adop ting those imp lementing acts concer ning Ar tif icial Inte lligence syste ms which are saf ety comp onents within

− the meaning of Regulation (EU) 2024/1689, the requirements set out in Chapt er III, Section 2, of that Regulation shall

− be take n into account.’;

− 120/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− (6) in Ar ticle 58, the f ollowing paragraph is added:

− ‘3. When adop ting deleg ated acts pursuant to paragraphs 1 and 2 concer ning Ar tif icial Intel ligence syste ms which

− are saf ety comp onents within the meaning of Regulation (EU) 2024/1689, the requirements set out in Chapt er III,

− Section 2, of that Regulation shall be taken into account.’ .

− Ar ticle 109

− Amendment to Regulation (EU) 2019/2144

− In Ar ticle 11 of Regulation (EU) 2019/2144, the f ollowing paragraph is added:

− ‘3. When adop ting the imp lementing acts pursuant to paragraph 2, concer ning ar tif icial intellig ence systems which are

− saf ety comp onents within the meaning of Regulation (EU) 2024/1689 of the European Parliame nt and of the Council (*),

− the requirements set out in Chapt er III, Section 2, of that Regulation shall be take n into account.

− (*) Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 la ying down har monised

− r ules on ar tif icial intellig ence and amending Regulations (EC) No 300/2008, (EU) No 167/2013, (EU) No 168/2013,

− (EU) 2018/858, (EU) 2018/1139 and (EU) 2019/2144 and Directives 2014/90/EU, (EU) 2016/797 and (EU)

− 2020/1828 (Ar tificial Inte lligence A ct) (OJ L, 2024/1689, 12.7.2024, ELI: http://data.europa.eu/eli/reg/

− 2024/1689/oj).’ .

− Ar ticle 110

− Amendment to Directiv e (EU) 2020/1828

− In Annex I to Directive (EU) 2020/1828 of the European Parliament and of the Council (

− 58

− ), the f ollowing point is added:

− ‘(68) Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 la ying down har monised

− r ules on ar tificial inte lligence and amending Regulations (EC) No 300/2008, (EU) No 167/2013, (EU) No 168/2013,

− (EU) 2018/858, (EU) 2018/1139 and (EU) 2019/2144 and Directives 2014/90/EU, (EU) 2016/797 and (EU)

− 2020/1828 (Ar tif icial Inte lligence A ct) (OJ L, 2024/1689, 12.7.2024, ELI: http://data.europa.eu/eli/reg/

− 2024/1689/oj).’.

− Ar ticle 111

− AI systems already placed on the mark et or put into ser vice and general-pur pose AI models already placed on the

− marked

− 1. Without prejudice to the application of Ar ticle 5 as referred to in Ar ticle 113(3), point (a), AI syste ms which are

− compo nents of the larg e-scale IT systems established b y the leg al acts listed in Annex X that hav e been placed on the marke t

− or put into ser vice bef ore 2 Augu st 2027 shall be brought into compliance with this Regulation b y 31 December 2030.

− The requirements laid do wn in this Regulation shall be taken into account in the evaluation of each larg e-scale IT system

− established by the leg al acts listed in Annex X t o be under take n as provid ed f or in those lega l acts and where those lega l acts

− are replaced or amended.

− 2. Without prejudice to the application of Ar ticle 5 as refer red t o in Ar ticle 113(3), point (a), this Regulation shall apply

− to operators of high-r isk AI systems, other than the syste ms refe r red to in paragraph 1 of this Ar ticle, that hav e been placed

− on the mark et or put into ser vice bef ore 2 Au gust 2026, only if, as from that date, those syste ms are subject to significant

− ch anges in their designs. In any case, the providers and deplo y ers of high-r isk AI systems intended to be used by public

− author ities shall take the necessar y steps to comply with the requirements and obliga tions of this Regulation by 2 August

− 2030.

− 3. Providers of general-purpo se AI models that have been placed on the marke t bef ore 2 August 2025 shall take the

− necessar y st eps in order to comp ly with the obliga tions laid do wn in this Regulation by 2 Augu st 2027.

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 121/144

− (

− 58

− ) Directive (EU) 2020/1828 of the European Parliament and of the Council of 25 No vember 2020 on representative actions f or the

− prot ection of the collective interests of consumers and repealing Directive 2009/22/EC (OJ L 409, 4.12.2020, p. 1).

− Ar ticle 112

− Evaluation and review

− 1. The Commission shall assess the need f or amendment of the list set out in Annex III and of the list of prohibite d AI

− practices laid do wn in Ar ticle 5, once a year f ollowing the entr y into f orce of this Regulation, and until the end of the per iod

− of the deleg ation of power laid down in Ar ticle 97. The Commission shall submit the findi ngs of that assessment t o the

− European Parliament and the Council.

− 2. By 2 August 2028 and ever y f our y ears thereaf ter , the Commission shall evaluate and repor t t o the European

− Parliame nt and to the Council on the f ollowing:

− (a) the need f or amendments extending existing area headings or adding new area headings in Annex III;

− (b) amendments to the list of AI systems requir ing additional transparency measures in Ar ticle 50;

− (c) amendments enhancing the effe ctiveness of the super vision and gove r nance system.

− 3. By 2 Au gust 2029 and ever y f our years thereaf ter , the Commission shall submit a repor t on the evaluation and review

− of this Regulation to the European Parliament and t o the Council. The repor t shall include an assessment with regard to the

− str ucture of enf orcement and the possible need f or a Union age ncy to resolve any identified shor tcomings. On the basis of

− the find ings, that repor t shall, where appropr iate, be accom panied by a proposal f or amendment of this Regulation. The

− repor ts shall be made public.

− 4. The repor ts refer red to in paragraph 2 shall pa y specif ic attention t o the f ollowing:

− (a) the status of the fi nancial, tec hnical and human resources of the national comp et ent author ities in order to eff ectively

− perfo r m the tasks assigned to them under this Regulation;

− (b) the stat e of penalties, in par ticular administrative fine s as refer red to in Ar ticle 99(1), applied by Member Stat es f or

− infr ingements of this Regulation;

− (c) adop ted har monised standards and common specifications developed to suppor t this Regulation;

− (d) the number of under takings that ent er the market af te r the entr y into application of this Regulation, and how many of

− them are SMEs.

− 5. By 2 August 2028, the Commission shall evaluate the functioning of the AI Offi ce, whether the AI Office has been

− given sufficient powers and comp etences to fulf il its tasks, and whether it would be relevant and needed f or the proper

− imp lementation and enf orcement of this Regulation to upgrade the AI Offi ce and its enf orcement compet ences and to

− increase its resources. The Commission shall submit a repor t on its evaluation to the European Parliame nt and to the

− Council.

− 6. By 2 Augu st 2028 and ever y f our y ears thereaf te r , the Commission shall submit a repor t on the review of the progress

− on the development of standardisation deliverables on the energy-eff icient development of g eneral-pur pose AI models, and

− asses the need f or fur ther measures or actions, including binding measures or actions. The repor t shall be submitted to the

− European Parliament and t o the Council, and it shall be made public.

− 7. By 2 Au gust 2028 and ever y three y ears thereaf ter , the Commission shall evaluate the imp act and effe ctiveness of

− v oluntar y codes of conduct to f oste r the application of the requirements set out in Chapt er III, Section 2 f or AI syste ms

− other than high-r isk AI syste ms and possibly other additional requirements f or AI syste ms other than high-r isk AI systems,

− including as regards environmental sustainability .

− 8. For the pur poses of paragraphs 1 to 7, the Board, the Member Stat es and national compet ent author ities shall provide

− the Commission with inf or mation upon its request and without undue dela y .

− 9. In car r ying out the evaluations and reviews refer red to in paragraphs 1 to 7, the Commission shall take into account

− the positions and find ings of the Board, of the European Parliame nt, of the Council, and of other relevant bodies or sources.

− 122/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− 10. The Commission shall, if necessar y , submit appropr iat e proposals to amend this Regulation, in par ticular taking into

− account developments in t echnology , the effect of AI syste ms on health and safety , and on fundamental r ights, and in light

− of the state of progress in the inf or mation society .

− 11. T o guide the evaluations and reviews referred t o in paragraphs 1 to 7 of this Ar ticle, the AI Office shall under take to

− develop an objective and par ticipative methodology f or the evaluation of r isk levels based on the cr ite r ia outlined in the

− relevant Ar ticles and the inclusion of new syste ms in:

− (a) the list set out in Annex III, including the extension of existing area headings or the addition of new area headings in

− that Annex;

− (b) the list of prohibited practices set out in Ar ticle 5; and

− (c) the list of AI syste ms requir ing additional transparency measures pursuant to Ar ticle 50.

− 12. Any amendment to this Regulation pursuant to paragraph 10, or relevant deleg ated or imp lementing acts, which

− concer ns sectoral Uni on har monisation legislation listed in Section B of Annex I shall take into account the regulator y

− specific ities of each secto r , and the existing gover nance, conf or mity assessment and enf orcement mechanisms and

− author ities established therein.

− 13. By 2 Au gust 2031, the Commission shall car r y out an assessment of the enf orcement of this Regulation and shall

− repor t on it t o the European Pa rliament, the Council and the European Economic and Social Committee, taking into

− account the first y ears of application of this Regulation. On the basis of the fi ndings, that repor t shall, where appropr iate, be

− accom panied by a proposal f or amendment of this Regulation with regard to the str ucture of enf orcement and the need f or

− a Uni on age ncy to resolve any identifie d shor tcomings.

− Ar ticle 113

− Entr y into force and application

− This Regulation shall ent er into f orce on the twentieth da y f ollowing that of its publication in the Of ficial Jour nal of the

− Europe an Union.

− It shall apply from 2 August 2026.

− How ever:

− (a) Chapt ers I and II shall apply from 2 Febr uar y 2025;

− (b) Chapt er III Section 4, Chap ter V , Chapt er VII and Chap ter XII and Ar ticle 78 shall apply from 2 Augu st 2025, with the

− ex ception of Ar ticle 101;

− (c) Ar ticle 6(1) and the cor responding obligations in this Regulation shall apply from 2 August 2027.

− This Regulation shall be binding in its entirety and directly applicable in all Member Stat es.

− Done at Br ussels, 13 June 2024.

− For the Eur opean P arliament

− The Pr esident

− R. METSOL A

− For the Council

− The Pr esident

− M. MICHEL

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 123/144

− ANNEX I

− Lis t of Union har monisation legislation

− Section A . List of Uni on har monisation legislation based on the New Legislative Framewo rk

− 1. Directive 2006/42/EC of the European Parliament and of the Council of 17 Ma y 2006 on machi ner y , and amending

− Directive 95/16/EC (OJ L 157, 9.6.2006, p. 24);

− 2. Directive 2009/48/EC of the European Pa rliament and of the Council of 18 June 2009 on the safety of to ys (OJ

− L 170, 30.6.2009, p. 1);

− 3. Directive 2013/53/EU of the European Parliame nt and of the Council of 20 November 2013 on recreational craf t

− and personal wat ercraf t and repealing Directive 94/25/EC (OJ L 354, 28.12.2013, p. 90);

− 4. Directive 2014/33/EU of the European Parliament and of the Council of 26 Febr uar y 2014 on the har monisation of

− the la ws of the Member States relating t o lif ts and saf ety comp onents f or lif ts (OJ L 96, 29.3.2014, p. 251);

− 5. Directive 2014/34/EU of the European Parliament and of the Council of 26 Febr uar y 2014 on the har monisation of

− the la ws of the Member States relating to equipment and protect ive systems intende d f or use in pote ntially explosive

− atmospheres (OJ L 96, 29.3.2014, p. 309);

− 6. Directive 2014/53/EU of the European Parliament and of the Council of 16 Apr il 2014 on the har monisation of the

− law s of the Member States relating to the making ava ilable on the marke t of radio equipment and repealing Directive

− 1999/5/EC (OJ L 153, 22.5.2014, p. 62);

− 7. Directive 2014/68/EU of the European Pa rliament and of the Council of 15 Ma y 2014 on the har monisation of the

− law s of the Member States relating to the making available on the marke t of pressure equipment (OJ L 189,

− 27.6.2014, p. 164);

− 8. Regulation (EU) 2016/424 of the European Parliament and of the Council of 9 Marc h 2016 on cablewa y

− installations and repealing Directive 2000/9/EC (OJ L 81, 31.3.2016, p. 1);

− 9. Regulation (EU) 2016/425 of the European Pa rliament and of the Council of 9 March 2016 on personal prot ective

− equipment and repealing Council Directive 89/686/EEC (OJ L 81, 31.3.2016, p. 51);

− 10. Regulation (EU) 2016/426 of the European Parliament and of the Council of 9 March 2016 on appliances bur ning

− g aseous fuels and repealing Directive 2009/142/EC (OJ L 81, 31.3.2016, p. 99);

− 11. Regulation (EU) 2017/745 of the European Pa rliament and of the Council of 5 Apr il 2017 on medical devices,

− amending Directive 2001/83/EC, Regulation (EC) No 178/2002 and Regulation (EC) No 1223/2009 and repealing

− Council Directives 90/385/EEC and 93/42/EEC (OJ L 117, 5.5.2017, p. 1);

− 12. Regulation (EU) 2017/746 of the European Parliament and of the Council of 5 Apr il 2017 on in vitro diagnostic

− medical devices and repealing Directive 98/79/EC and Commission Decision 2010/227/EU (OJ L 117, 5.5.2017,

− p. 176).

− Section B. List of other Uni on har monisation legislation

− 13. Regulation (EC) No 300/2008 of the European Parliame nt and of the Council of 11 March 2008 on common r ules

− in the fie ld of civil a viation secur ity and repealing Regulation (EC) No 2320/2002 (OJ L 97, 9.4.2008, p. 72);

− 14. Regulation (EU) No 168/2013 of the European Pa rliament and of the Council of 15 Januar y 2013 on the approval

− and marke t sur veillance of two- or three-wheel vehicles and quadr icy cles (OJ L 60, 2.3.2013, p. 52);

− 15. Regulation (EU) No 167/2013 of the European Parliament and of the Council of 5 Febr uar y 2013 on the approval

− and marke t sur veillance of agr icultural and f orestr y vehicles (OJ L 60, 2.3.2013, p. 1);

− 124/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− 16. Directive 2014/90/EU of the European Parliament and of the Council of 23 July 2014 on mar ine equipment and

− repealing Council Directive 96/98/EC (OJ L 257, 28.8.2014, p. 146);

− 17. Directive (EU) 2016/797 of the European Parliament and of the Council of 11 Ma y 2016 on the interop erability of

− the rail syste m within the European Union (OJ L 138, 26.5.2016, p. 44);

− 18. Regulation (EU) 2018/858 of the European Parliament and of the Council of 30 Ma y 2018 on the approval and

− marke t sur veillance of motor vehicles and their trailers, and of systems, compo nents and separate tec hnical units

− intende d f or such vehicles, amending Regulations (EC) No 715/2007 and (EC) No 595/2009 and repealing Directive

− 2007/46/EC (OJ L 151, 14.6.2018, p. 1);

− 19. Regulation (EU) 2019/2144 of the European Parliament and of the Council of 27 No vember 2019 on type-appro val

− requirements f or motor vehicles and their trailers, and syste ms, compo nents and separate tec hnical units intended

− f or suc h vehicles, as rega rds their g eneral safety and the prot ection of vehicle occupants and vulnerable road users,

− amending Regulation (EU) 2018/858 of the European Parliament and of the Council and repealing Regulations (EC)

− No 78/2009, (EC) No 79/2009 and (EC) No 661/2009 of the European Parliament and of the Council and

− Commission Regulations (EC) No 631/2009, (EU) No 406/2010, (EU) No 672/2010, (EU) No 1003/2010,

− (EU) No 1005/2010, (EU) No 1008/2010, (EU) No 1009/2010, (EU) No 19/2011, (EU) No 109/2011, (EU)

− No 458/2011, (EU) No 65/2012, (EU) No 130/2012, (EU) No 347/2012, (EU) No 351/2012, (EU) No 1230/2012

− and (EU) 2015/166 (OJ L 325, 16.12.2019, p. 1);

− 20. Regulation (EU) 2018/1139 of the European Parliament and of the Council of 4 July 2018 on common r ules in the

− fie ld of civil aviat ion and establishing a European Uni on A viation Saf ety Ag ency , and amending Regulations (EC)

− No 2111/2005, (EC) No 1008/2008, (EU) No 996/2010, (EU) No 376/2014 and Directives 2014/30/EU and

− 2014/53/EU of the European Pa rliament and of the Council, and repealing Regulations (EC) No 552/2004 and (EC)

− No 216/2008 of the European Pa rliament and of the Council and Council Regulation (EEC) No 3922/91 (OJ L 212,

− 22.8.2018, p. 1), in so f ar as the design, production and placing on the market of aircraf ts refe r red to in Ar ticle 2(1),

− points (a) and (b) thereof, where it concer ns unmanned aircraf t and their engines, propellers, par ts and equipment to

− control them remotely , are concer ned.

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 125/144

− ANNEX II

− Lis t of cr iminal offences referred to in Ar ticle 5(1), f irst subparag raph, point (h)(iii)

− Cr iminal offences refe r red t o in Ar ticle 5(1), fir st subparagraph, point (h)(iii):

− — t er ror ism,

− — trafficking in human beings,

− — sexual exploitation of children, and child por nography ,

− — illicit traffi ck ing in narcotic dr ugs or psy ch otropic substances,

− — illicit traffi ck ing in weapons, munitions or explosives,

− — murder , gr ievous bodily injur y ,

− — illicit trade in human org ans or tissue,

− — illicit traffi ck ing in nuclear or radioactive materi als,

− — kidnapping, illegal restraint or hostage -taking,

− — cr imes within the jur isdiction of the Inte r national Cr iminal Cour t,

− — unla wful seizure of aircraf t or ships,

− — rape,

− — en vironmental cr ime,

− — org anised or ar med robber y ,

− — sabotag e,

− — par ticipation in a cr iminal org anisation inv olved in one or more of the offences listed above .

− 126/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− ANNEX III

− High-r isk AI sys tems referred to in Ar ticle 6(2)

− High-r isk AI systems pursuant to Ar ticle 6(2) are the AI systems listed in an y of the f ollo wing areas:

− 1. Biometr ics, in so f ar as their use is per mitted under relevant Union or national law:

− (a) remote biometr ic identifica tion syste ms.

− This shall not include AI systems intende d to be used f or biometr ic ver ifi cation the sole pur pose of which is to

− confir m that a specific natural person is the person he or she claims to be;

− (b) AI systems inte nded to be used f or biometr ic cate gor isation, according to sensitive or protect ed attr ibutes or

− ch aracter istics based on the inference of those attr ibut es or ch aracteristics ;

− (c) AI systems inte nded t o be used f or emotion recognition.

− 2. Cr itical infrastr ucture: AI systems intended to be used as safety comp onents in the managem ent and operation of

− cr itical digital infrastr ucture, road traff ic, or in the supply of wate r , g as, heating or electr icity .

− 3. Education and v ocational training:

− (a) AI syste ms intended to be used to determine access or admission or to assign natural persons to educational and

− v ocational training institutions at all levels;

− (b) AI syste ms inte nded to be used to evaluate lear ning outcomes, including when those outcomes are used to steer

− the lear ning process of natural persons in educational and v ocational training institutions at all levels;

− (c) AI syste ms inte nded to be used f or the pur pose of assessing the appropr iate level of education that an individual

− will receive or will be able to access, in the context of or within educational and v ocational training institutions

− at all levels;

− (d) AI systems inte nded to be used f or monitoring and det ecting prohibite d behavi our of students dur ing te sts in the

− cont ext of or within educational and vocational training institutions at all levels.

− 4. Emplo yment, wo rkers’ managem ent and access to self-emplo yment:

− (a) AI syste ms intended to be used f or the recr uitment or selection of natural persons, in par ticular t o place targe ted

− job adver tisements, to analyse and filter job applications, and to evaluate candidates;

− (b) AI systems intended to be used to mak e decisions aff ecting te r ms of work -relate d relationships, the promotion or

− te r mination of wo rk -related contractual relationships, t o allocat e tasks based on individual behavi our or personal

− traits or ch aracter istics or to monito r and evaluate the perf or mance and behavi our of persons in such

− relationships.

− 5. A ccess to and enjo yment of essential pr ivat e ser vices and essential public ser vices and benefits:

− (a) AI syste ms intended to be used b y public author ities or on behalf of public author ities t o evaluate the eligibility

− of natural persons f or essential public assistance benefits and ser vices, including healthcare ser vices, as well as to

− grant, reduce, revok e, or reclaim such benefi ts and ser vices;

− (b) AI syste ms intended to be used to evaluate the creditwo r thiness of natural persons or establish their credit score,

− with the ex ception of AI systems used f or the pur pose of detect ing financ ial fraud;

− (c) AI syste ms intended to be used f or r isk assessment and pr icing in relation to natural persons in the case of life

− and health insurance;

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 127/144

− (d) AI systems inte nded to evaluate and classify emerge ncy calls by natural persons or to be used to dispatch , or to

− establish pr ior ity in the dispatching of, emergency fir st response ser vices, including b y police, firef ighters and

− medical aid, as well as of emerg ency healthcare patient tr iage systems.

− 6. Law enf orcement, in so f ar as their use is per mitted under relevant Union or national law:

− (a) AI syste ms intended to be used by or on behalf of law enf orcement author ities, or b y Uni on institutions, bodies,

− off ices or age ncies in suppor t of law enf orcement author ities or on their behalf to assess the r isk of a natural

− person becoming the victim of cr iminal offences;

− (b) AI syste ms intended to be used by or on behalf of law enf orcement author ities or by Uni on institutions, bodies,

− off ices or age ncies in suppor t of la w enf orcement author ities as polygraphs or similar tools;

− (c) AI syste ms intended to be used by or on behalf of law enf orcement author ities, or b y Uni on institutions, bodies,

− off ices or age ncies, in suppor t of la w enf orcement author ities t o evaluate the reliability of evidence in the course

− of the invest igation or prosecution of cr iminal offences;

− (d) AI syste ms intende d to be used by law enf orcement author ities or on their behalf or b y Uni on institutions,

− bodies, offices or age ncies in suppor t of law enf orcement author ities f or assessing the r isk of a natural person

− offending or re-offending not solely on the basis of the profil ing of natural persons as referred to in Ar ticle 3(4)

− of Directive (EU) 2016/680, or to assess personality traits and ch aracter istics or past cr iminal behavi our of

− natural persons or groups;

− (e) AI syste ms intended to be used by or on behalf of law enf orcement author ities or by Uni on institutions, bodies,

− off ices or age ncies in suppor t of law enf orcement author ities f or the profil ing of natural persons as refer red to in

− Ar ticle 3(4) of Directive (EU) 2016/680 in the course of the detection, invest igation or prosecution of cr iminal

− offenc es.

− 7. Migration, asylum and border control management, in so f ar as their use is per mitte d under relevant Uni on or

− national law:

− (a) AI systems intended to be used by or on behalf of compet ent public author ities or by Uni on institutions, bodies,

− off ices or age ncies as polygraphs or similar too ls;

− (b) AI systems intended to be used by or on behalf of compet ent public author ities or by Uni on institutions, bodies,

− off ices or agencies to assess a r isk, including a secur ity r isk, a r isk of ir regular migration, or a health r isk, posed

− by a natural person who inte nds to ent er or who has entered into the terr itory of a Member State;

− (c) AI systems intended to be used by or on behalf of compet ent public author ities or by Uni on institutions, bodies,

− off ices or agencies to assist compet ent public author ities f or the examination of applications f or asylum, visa or

− residence per mits and f or associate d complaints with rega rd to the eligibility of the natural persons applying f or

− a status, including relate d assessments of the reliability of evidence;

− (d) AI systems intended to be used by or on behalf of comp etent public author ities, or by Uni on institutions, bodies,

− off ices or agencies, in the cont ext of migration, asylum or border control management, f or the pur pose of

− detect ing, recognising or identifying natural persons, with the ex ception of the ver ificati on of trave l documents.

− 8. Ad ministration of justice and democratic processes:

− (a) AI syste ms intended to be used by a judicial author ity or on their behalf to assist a judicial author ity in

− researchi ng and inter preting f acts and the law and in applying the law to a concrete set of f acts, or to be used in

− a similar wa y in alt er native dispute resolution;

− 128/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− (b) AI syste ms intended t o be used f or inf luencing the outcome of an election or refere ndum or the v oting

− behavio ur of natural persons in the ex ercise of their vot e in elections or refe renda. This does not include AI

− syste ms to the output of which natural persons are not directly exposed, such as too ls used t o org anise, opti mise

− or str ucture political camp aigns from an administrative or logistical point of view .

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 129/144

− ANNEX IV

− T echnical document ation referred to in Ar ticle 11(1)

− The te chnical documentation refe r red to in Ar ticle 11(1) shall contain at least the f ollowing inf or mation, as applicable to

− the relevant AI syste m:

− 1. A g eneral descr iption of the AI syste m including:

− (a) its inte nded pur pose, the name of the provider and the version of the syste m ref lecting its relation t o previous

− versions;

− (b) ho w the AI system inte racts with, or can be used to interac t with, hardware or sof tware, including with other AI

− syste ms, that are not par t of the AI syste m itself, where applicable;

− (c) the versions of relevant sof tware or fir m ware, and any requirements related to version update s;

− (d) the descr ipt ion of all the f or ms in which the AI system is placed on the marke t or put into ser vice, suc h as

− sof tware pack ages embedded into hardware, do wnloads, or APIs;

− (e) the descr ipt ion of the hardware on which the AI system is intended to r un;

− (f) where the AI system is a comp onent of products, photographs or illustrations showing exte r nal f eatures, the

− marking and inte r nal la y out of those products;

− (g) a basic descr ipt ion of the user -interface provided t o the deplo yer;

− (h) instr uctions f or use f or the deplo y er , and a basic descr ipt ion of the user -inte rface pro vided to the deplo yer , where

− applicable;

− 2. A detailed descr iption of the elements of the AI syste m and of the process f or its development, including:

− (a) the methods and st eps perfo r med f or the development of the AI syste m, including, where relevant, recourse to

− pre-trained systems or tools pro vided by third par ties and how those were used, integrat ed or modifi ed b y the

− provid er ;

− (b) the design specif ications of the syste m, namely the g eneral logic of the AI system and of the algor ithms; the k ey

− design choices including the rationale and assump tions made, including with regard to persons or groups of

− persons in respect of who, the system is intended to be used; the main classific ation ch oices; what the syste m is

− designed to op timise f or , and the relevance of the different parameters; the descr iption of the expected output

− and output quality of the system; the decisions about any possible trade-off made regard ing the t echnical

− solutions adopt ed t o comply with the requirements set out in Chapt er III, Section 2;

− (c) the descr iption of the system arch itecture explaining how sof tware comp onents build on or feed into each other

− and integrat e into the overall processing; the comp utational resources used to develop, train, test and validat e the

− AI system;

− (d) where relevant, the data requirements in t er ms of datasheets descr ibing the training methodologies and

− te chniques and the training data sets used, including a g eneral descr iption of these data sets, inf or mation about

− their prove nance, scope and main ch aracter istics; ho w the data was obtained and selected; labelling procedures

− (e.g. f or super vised lear ning), data cleaning methodologies (e.g. outliers detection);

− (e) assessment of the human o versight measures needed in accordance with Ar ticle 14, including an assessment of

− the te ch nical measures needed to f acilitate the inte r pretation of the outputs of AI syste ms by the deplo yers, in

− accordance with Ar ticle 13(3), point (d);

− (f) where applicable, a detailed descr ipt ion of pre-deter mined chang es to the AI system and its perfo r mance,

− tog ether with all the relevant inf or mation relate d to the te chnical solutions adop te d to ensure continuous

− complia nce of the AI system with the relevant requirements set out in Chapt er III, Section 2;

− (g) the validation and te sting procedures used, including inf or mation about the validation and te sting data used and

− their main ch aracteristics ; metr ics used t o measure accuracy , robustness and compliance with other relevant

− requirements set out in Chap ter III, Section 2, as well as pote ntially discr iminat or y imp acts; test logs and all t est

− repor ts dated and signed by the responsible persons, including with regard to pre-determ ined ch anges as refe r red

− to under point (f);

− 130/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− (h) cybersecur ity measures put in place;

− 3. Detailed inf or mation about the monitoring, functioning and control of the AI syste m, in par ticular with regard to :

− its capabilities and limitations in perf or mance, including the degrees of accuracy f or specif ic persons or groups of

− persons on which the system is intended to be used and the overall expected level of accuracy in relation to its

− intende d pur pose; the f oreseeable unintended outcomes and sources of r isks t o health and safety , fundamental r ights

− and discr imination in view of the intended pur pose of the AI system; the human oversight measures needed in

− accordance with Ar ticle 14, including the te chnical measures put in place to f acilitate the inte r pretation of the

− outputs of AI syste ms by the deplo yers; specifications on input data, as appropr iate;

− 4. A descr iption of the appropr iat eness of the perf or mance metr ics f or the specific AI syste m;

− 5. A detailed descr iption of the r isk managem ent syste m in accordance with Ar ticle 9;

− 6. A descr iption of relevant ch anges made by the provid er to the syste m through its lif ecy cle;

− 7. A list of the har monised standards applied in full or in par t the refere nces of which hav e been published in the

− Off icial Jour nal of the European Union; where no suc h har monised standards hav e been applied, a detailed descr iption

− of the solutions adop ted to meet the requirements set out in Chapt er III, Section 2, including a list of other relevant

− standards and te chnical specifications applied;

− 8. A cop y of the EU declaration of conf or mity refer red to in Ar ticle 47;

− 9. A detailed descr iption of the system in place to evaluate the AI system perf or mance in the post-mark et phase in

− accordance with Ar ticle 72, including the post-mark et monito r ing plan refer red to in Ar ticle 72(3).

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 131/144

− ANNEX V

− EU declaration of confor mity

− The EU declaration of conf or mity refer red to in Ar ticle 47, shall contain all of the f ollowi ng inf or mation:

− 1. AI syste m name and type and any additional unambiguous refe rence allowi ng the identifi cation and traceability of

− the AI syste m;

− 2. The name and address of the provid er or , where applicable, of their author ised representative;

− 3. A statem ent that the EU declaration of conf or mity referred t o in Ar ticle 47 is issued under the sole responsibility of

− the provid er ;

− 4. A stat ement that the AI syste m is in conf or mity with this Regulation and, if applicable, with any other relevant

− Uni on law that provides f or the issuing of the EU declaration of conf or mity refe r red to in Ar ticle 47;

− 5. Where an AI syste m inv olves the processing of personal data, a statement that that AI syste m comp lies with

− Regulations (EU) 2016/679 and (EU) 2018/1725 and Directive (EU) 2016/680;

− 6. Ref erences to any relevant har monised standards used or any other common specification in relation to which

− conf or mity is declared;

− 7. Where applicable, the name and identifica tion number of the notified body , a descr iption of the conf or mity

− assessment procedure perf or med, and identification of the cer tif icat e issued;

− 8. The place and date of issue of the declaration, the name and function of the person who signed it, as well as an

− indication f or , or on behalf of whom, that person signed, a signature.

− 132/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− ANNEX VI

− Conform ity assessment procedure based on inter nal control

− 1. The conf or mity assessment procedure based on inte r nal control is the conf or mity assessment procedure based on

− points 2, 3 and 4.

− 2. The pro vider ver ifi es that the established quality management system is in comp liance with the requirements of

− Ar ticle 17.

− 3. The provider examines the inf or mation contained in the te chnical documentation in order to assess the compliance

− of the AI system with the relevant essential requirements set out in Chapt er III, Section 2.

− 4. The provider also ver ifies that the design and development process of the AI system and its post-mark et monitoring

− as refer red to in Ar ticle 72 is consiste nt with the te chnical documentation.

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 133/144

− ANNEX VII

− Conf or mity based on an assessment of the quality management system and an assessment of the

− technical documentation

− 1. Introduction

− Conf or mity based on an assessment of the quality management syste m and an assessment of the t echnical

− documentation is the conf or mity assessment procedure based on points 2 to 5.

− 2. Over view

− The approved quality managem ent syste m f or the design, development and t esting of AI systems pursuant to

− Ar ticle 17 shall be examined in accordance with point 3 and shall be subject to sur veillance as specified in point 5.

− The te chnical documentation of the AI system shall be examined in accordance with point 4.

− 3. Quality managem ent syste m

− 3.1. The application of the provider shall include:

− (a) the name and address of the provid er and, if the application is lodged by an author ised representative, also their

− name and address;

− (b) the list of AI syste ms covered under the same quality managem ent syste m;

− (c) the t echnical documentation f or each AI system covered under the same quality managem ent syste m;

− (d) the documentation concer ning the quality managem ent syste m which shall co ver all the aspects listed under

− Ar ticle 17;

− (e) a descr iption of the procedures in place to ensure that the quality management system remains adequate and

− effe ctive;

− (f) a wr itten declaration that the same application has not been lodg ed with any other notified body .

− 3.2. The quality managem ent system shall be assessed b y the notified body , which shall determine whether it satisfies the

− requirements refe r red to in Ar ticle 17.

− The decision shall be notif ied to the provider or its author ised representative.

− The notification shall contain the conclusions of the assessment of the quality management syste m and the reasoned

− assessment decision.

− 3.3. The quality manag ement syste m as approved shall continue to be implement ed and maintained by the provid er so

− that it remains adequate and eff icient.

− 3.4. Any intended chang e to the approve d quality manag ement system or the list of AI syste ms covered by the latter shall

− be brought to the attention of the notified body by the pro vider .

− The proposed ch anges shall be examined by the notified body , which shall decide whether the modif ied quality

− managem ent system continues to satisfy the requirements referred to in point 3.2 or whether a reassessment is

− necessar y .

− The notif ied body shall notify the provider of its decision. The notif ication shall contain the conclusions of the

− examination of the ch anges and the reasoned assessment decision.

− 4. Control of the t echnical documentation.

− 4.1. In addition to the application refer red to in point 3, an application with a notif ied body of their choice shall be

− lodg ed by the provider f or the assessment of the te chnical documentation relating to the AI system which the

− provid er inte nds to place on the market or put into ser vice and which is covered by the quality managem ent system

− referred to under point 3.

− 4.2. The application shall include:

− (a) the name and address of the provider;

− (b) a wr itten declaration that the same application has not been lodg ed with any other notified body ;

− (c) the t echnical documentation refer red to in Annex IV .

− 134/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− 4.3. The tec hnical documentation shall be examined b y the notified body . Where relevant, and limited to what is

− necessar y to fulf il its tasks, the notif ied body shall be granted full access t o the training, validation, and te sting data

− sets used, including, where appropr iate and subject to secur ity safegua rds, through API or other relevant t echnical

− means and too ls enabling remote access.

− 4.4. In examining the te ch nical documentation, the notif ied body ma y require that the provider supply fur ther evidence

− or car r y out fur ther te sts so as to enable a proper assessment of the conf or mity of the AI syste m with the

− requirements set out in Chapt er III, Section 2. Where the notif ied body is not satisf ied with the te sts car r ied out by

− the provid er , the notif ied body shall itself directly car r y out adequat e te sts, as appropr iate.

− 4.5. Where necessar y to assess the conf or mity of the high-r isk AI syste m with the requirements set out in Chapt er III,

− Section 2, af ter all other reasonable means to ver ify conf or mity hav e been exhaust ed and hav e pro ven to be

− insuffic ient, and upon a reasoned request, the notif ied body shall also be grante d access to the training and trained

− models of the AI system, including its relevant parameters. Such access shall be subject t o existing Union law on the

− protect ion of intellectual proper ty and trade secrets.

− 4.6. The decision of the notif ied body shall be notif ied t o the provider or its author ised representative. The notif ication

− shall contain the conclusions of the assessment of the t echnical documentation and the reasoned assessment

− decision.

− Where the AI system is in conf or mity with the requirements set out in Chapt er III, Section 2, the notif ied body shall

− issue a Union te chnical documentation assessment cer tif icat e. The cer tif icat e shall indicate the name and address of

− the provider , the conclusions of the examination, the conditions (if any) f or its validity and the data necessar y f or the

− identifica tion of the AI syste m.

− The cer tif icat e and its annexe s shall contain all relevant inf or mation to allow the conf or mity of the AI system to be

− evaluated, and to allow f or control of the AI system while in use, where applicable.

− Where the AI syste m is not in conf or mity with the requirements set out in Chap ter III, Section 2, the notif ied body

− shall refuse to issue a Union te ch nical documentation assessment cer tif icat e and shall inf or m the applicant

− according ly , giving detailed reasons f or its refusal.

− Where the AI syste m does not meet the requirement relating to the data used t o train it, re-training of the AI system

− will be needed pr ior to the application f or a new conf or mity assessment. In this case, the reasoned assessment

− decision of the notif ied body refusing to issue the Uni on te chnical documentation assessment cer tif icat e shall

− contain specific considerations on the quality data used to train the AI syste m, in par ticular on the reasons f or

− non-comp liance.

− 4.7. Any chang e to the AI system that could affe ct the comp liance of the AI system with the requirements or its intended

− pur pose shall be assessed by the notified body which issued the Uni on te chnical documentation assessment

− cer tif icat e. The provider shall inf or m such notif ied body of its intention to introduce any of the abovementioned

− ch anges, or if it other wise becomes aw are of the occur rence of suc h ch anges. The inte nded ch anges shall be assessed

− by the notif ied body , which shall decide whether those ch anges require a new conf or mity assessment in accordance

− with Ar ticle 43(4) or whether they could be addressed by means of a supplement t o the Uni on t echnical

− documentation assessment cer tif icat e. In the latter case, the notif ied body shall assess the chang es, notify the

− provid er of its decision and, where the ch ange s are approve d, issue to the provider a supplement to the Uni on

− te chnical documentation assessment cer tif icate.

− 5. Sur veillance of the approve d quality managem ent syste m.

− 5.1. The pur pose of the sur veillance car r ied out by the notified body referred to in P oint 3 is to mak e sure that the

− provid er duly complie s with the te r ms and conditions of the approve d quality managem ent syste m.

− 5.2. For assessment pur poses, the provider shall allow the notif ied body t o access the premises where the design,

− development, testing of the AI systems is taking place. The provid er shall fur ther share with the notif ied body all

− necessar y inf or mation.

− 5.3. The notified body shall car r y out per iodic audits to mak e sure that the provid er maintains and applies the quality

− managem ent system and shall provide the provid er with an audit repor t. In the context of those audits, the notif ied

− body ma y car r y out additional tests of the AI syste ms f or which a Uni on te ch nical documentation assessment

− cer tif icat e was issued.

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 135/144

− ANNEX VIII

− Infor mation to be submitted upon the registration of high-r isk AI systems in accordance with

− Ar ticle 49

− Section A — Inf or mation to be submitted by provid ers of high-r isk AI syste ms in accordance with Ar ticle 49(1)

− The f ollowing inf or mation shall be provid ed and thereaf te r k ept up to date with rega rd to high-r isk AI systems t o be

− register ed in accordance with Ar ticle 49(1):

− 1. The name, address and contact details of the provid er ;

− 2. Where submission of inf or mation is car r ied out by another person on behalf of the provider , the name, address and

− contact details of that person;

− 3. The name, address and contact details of the author ised representative, where applicable;

− 4. The AI system trade name and any additional unambiguous refe rence allo wing the identification and traceability of

− the AI syste m;

− 5. A descr iption of the inte nded pur pose of the AI system and of the comp onents and functions suppor te d through

− this AI system;

− 6. A basic and concise descr iption of the inf or mation used by the syste m (data, inputs) and its operating logic;

− 7. The status of the AI system (on the mark et, or in ser vice; no longer placed on the market/in ser vice, recalled);

− 8. The type, number and expir y date of the cer tif icate issued b y the notif ied body and the name or identification

− number of that notif ied body , where applicable;

− 9. A scanned cop y of the cer tif icat e referred to in point 8, where applicable;

− 10. Any Member Stat es in which the AI syste m has been placed on the marke t, put into ser vice or made ava ilable in the

− Uni on;

− 11. A cop y of the EU declaration of conf or mity refer red to in Ar ticle 47;

− 12. Electronic instr uctions f or use; this inf or mation shall not be provid ed f or high-r isk AI syste ms in the areas of law

− enf orcement or migration, asylum and border control management referred to in Annex III, points 1, 6 and 7;

− 13. A URL f or additional inf or mation (optional).

− Section B — Inf or mation to be submitt ed by provider s of high-r isk AI syste ms in accordance with Ar ticle 49(2)

− The f ollowing inf or mation shall be provided and thereaf ter k ept up to dat e with regard to AI syste ms t o be regist ered in

− accordance with Ar ticle 49(2):

− 1. The name, address and contact details of the provid er ;

− 2. Where submission of inf or mation is car r ied out by another person on behalf of the provider , the name, address and

− contact details of that person;

− 3. The name, address and contact details of the author ised representative, where applicable;

− 4. The AI system trade name and any additional unambiguous refe rence allo wing the identification and traceability of

− the AI syste m;

− 5. A descr iption of the intended pur pose of the AI system;

− 6. The condition or conditions under Ar ticle 6(3)based on which the AI syste m is considered to be not-high-r isk;

− 7. A shor t summar y of the grounds on which the AI system is considered to be not-high-r isk in application of the

− procedure under Ar ticle 6(3);

− 8. The status of the AI system (on the mark et, or in ser vice; no longer placed on the market/in ser vice, recalled);

− 9. Any Member Stat es in which the AI syste m has been placed on the marke t, put into ser vice or made ava ilable in the

− Uni on.

− 136/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− Section C — Inf or mation to be submitt ed by deplo y ers of high-r isk AI systems in accordance with Ar ticle 49(3)

− The f ollowing inf or mation shall be provid ed and thereaf te r k ept up to date with rega rd to high-r isk AI systems t o be

− register ed in accordance with Ar ticle 49(3):

− 1. The name, address and contact details of the deplo y er ;

− 2. The name, address and contact details of the person submitting inf or mation on behalf of the deplo y er ;

− 3. The URL of the entr y of the AI syste m in the EU database b y its provider;

− 4. A summar y of the fi ndings of the fundamental r ights impact assessment conducted in accordance with Ar ticle 27;

− 5. A summar y of the data protect ion imp act assessment car r ied out in accordance with Ar ticle 35 of Regulation (EU)

− 2016/679 or Ar ticle 27 of Directive (EU) 2016/680 as specified in Ar ticle 26(8) of this Regulation, where

− applicable.

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 137/144

− ANNEX IX

− Inf or mation t o be submitted upon the registration of high-r isk AI systems listed in Annex III in

− relation t o testi ng in real w orld conditions in accordance with Ar ticle 60

− The f ollowing inf or mation shall be provid ed and thereaf ter kep t up to date with regard to te sting in real world conditions to

− be registered in accordance with Ar ticle 60:

− 1. A Uni on-wide unique sing le identifica tion number of the te sting in real world conditions;

− 2. The name and contact details of the provider or prospective provid er and of the deplo y ers in v olved in the te sting in

− real w orld conditions;

− 3. A br ief descr iption of the AI syste m, its inte nded pur pose, and other inf or mation necessar y f or the identification of

− the syste m;

− 4. A summar y of the main charact er istics of the plan f or te sting in real world conditions;

− 5. Inf or mation on the suspension or te r mination of the te sting in real world conditions.

− 138/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− ANNEX X

− Union legislative acts on larg e-scale IT sys tems in the area of Freedom, Secur ity and Justice

− 1. Sc hengen Inf or mation System

− (a) Regulation (EU) 2018/1860 of the European Pa rliament and of the Council of 28 November 2018 on the use of

− the Sc hengen Inf or mation System f or the retur n of illegally sta ying third-countr y nationals (OJ L 312,

− 7.12.2018, p. 1).

− (b) Regulation (EU) 2018/1861 of the European Parliament and of the Council of 28 November 2018 on the

− establishment, operation and use of the Sc hengen Inf or mation System (SIS) in the fie ld of border ch ec k s, and

− amending the Convention im plementing the Scheng en Agreement, and amending and repealing Regulation (EC)

− No 1987/2006 (OJ L 312, 7.12.2018, p. 14).

− (c) Regulation (EU) 2018/1862 of the European Parliament and of the Council of 28 November 2018 on the

− establishment, operation and use of the Sche ngen Inf or mation Syste m (SIS) in the fie ld of police cooperation and

− judicial cooperation in cr iminal matt ers, amending and repealing Council Decision 2007/533/JHA, and

− repealing Regulation (EC) No 1986/2006 of the European Parliame nt and of the Council and Commission

− Decision 2010/261/EU (OJ L 312, 7.12.2018, p. 56).

− 2. Visa Inf or mation Syst em

− (a) Regulation (EU) 2021/1133 of the European Parliament and of the Council of 7 July 2021 amending

− Regulations (EU) No 603/2013, (EU) 2016/794, (EU) 2018/1862, (EU) 2019/816 and (EU) 2019/818 as regard s

− the establishment of the conditions f or accessing other EU inf or mation syste ms f or the pur poses of the Visa

− Inf or mation Syste m (OJ L 248, 13.7.2021, p. 1).

− (b) Regulation (EU) 2021/1134 of the European Parliament and of the Council of 7 July 2021 amending

− Regulations (EC) No 767/2008, (EC) No 810/2009, (EU) 2016/399, (EU) 2017/2226, (EU) 2018/1240, (EU)

− 2018/1860, (EU) 2018/1861, (EU) 2019/817 and (EU) 2019/1896 of the European Parliament and of the

− Council and repealing Council Decisions 2004/512/EC and 2008/633/JHA, f or the pur pose of ref or ming the

− Visa Inf or mation Syst em (OJ L 248, 13.7.2021, p. 11).

− 3. Eurodac

− Regulation (EU) 2024/1358 of the European Parliament and of the Council of 14 Ma y 2024 on the establishment of

− ‘Eurodac’ f or the comp ar ison of biometr ic data in order to eff ectively apply Regulations (EU) 2024/1315 and (EU)

− 2024/1350 of the European Parliament and of the Council and Council Directive 2001/55/EC and to identify

− illegally sta ying third-countr y nationals and stat eless persons and on requests f or the compari son with Eurodac data

− by Member Stat es’ law enf orcement author ities and Europol f or law enf orcement pur poses, amending Regulations

− (EU) 2018/1240 and (EU) 2019/818 of the European Pa rliament and of the Council and repealing Regulation (EU)

− No 603/2013 of the European Parliament and of the Council (OJ L, 2024/1358, 22.5.2024, ELI: http://data.europa.

− eu/eli/reg/2024/1358/oj).

− 4. Entr y/Exit System

− Regulation (EU) 2017/2226 of the European Pa rliament and of the Council of 30 November 2017 establishing an

− Entr y/Exit System (EES) t o regist er entr y and exit data and refusal of entr y data of third-countr y nationals crossing

− the exte r nal borders of the Member Stat es and determining the conditions f or access to the EES f or la w enf orcement

− pur poses, and amending the Conve ntion im plementing the Sc hengen Agreement and Regulations (EC)

− No 767/2008 and (EU) No 1077/2011 (OJ L 327, 9.12.2017, p. 20).

− 5. European T ravel Inf or mation and Author isation Syste m

− (a) Regulation (EU) 2018/1240 of the European Parliament and of the Council of 12 Sep tember 2018 establishing

− a European T ravel Inf or mation and Au thor isation Syst em (ETIAS) and amending Regulations (EU)

− No 1077/2011, (EU) No 515/2014, (EU) 2016/399, (EU) 2016/1624 and (EU) 2017/2226 (OJ L 236,

− 19.9.2018, p. 1).

− (b) Regulation (EU) 2018/1241 of the European Parliame nt and of the Council of 12 Sept ember 2018 amending

− Regulation (EU) 2016/794 f or the pur pose of establishing a European T rave l Inf or mation and Author isation

− Syste m (ETIAS) (OJ L 236, 19.9.2018, p. 72).

− OJ L, 12.7.2024

− EN

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 139/144

− 6. European Cr iminal Records Inf or mation Syst em on third-countr y nationals and stat eless persons

− Regulation (EU) 2019/816 of the European Parliament and of the Council of 17 Apr il 2019 establishing

− a centralised system f or the identifica tion of Member States holding conviction inf or mation on third-countr y

− nationals and statele ss persons (ECRIS- TCN) to supplement the European Cr iminal Records Inf or mation Syste m and

− amending Regulation (EU) 2018/1726 (OJ L 135, 22.5.2019, p. 1).

− 7. Interop erability

− (a) Regulation (EU) 2019/817 of the European Parliament and of the Council of 20 Ma y 2019 on establishing

− a framework f or interoperability between EU inf or mation syste ms in the fi eld of borders and visa and amending

− Regulations (EC) No 767/2008, (EU) 2016/399, (EU) 2017/2226, (EU) 2018/1240, (EU) 2018/1726 and (EU)

− 2018/1861 of the European Pa rliament and of the Council and Council Decisions 2004/512/EC and

− 2008/633/JHA (OJ L 135, 22.5.2019, p. 27).

− (b) Regulation (EU) 2019/818 of the European Parliament and of the Council of 20 Ma y 2019 on establishing

− a framework f or interoperability between EU inf or mation syste ms in the field of police and judicial cooperation,

− asylum and migration and amending Regulations (EU) 2018/1726, (EU) 2018/1862 and (EU) 2019/816 (OJ

− L 135, 22.5.2019, p. 85).

− EN

− OJ L, 12.7.2024

− 140/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− ANNEX XI

− T echnical document ation referred to in Ar ticle 53(1), point (a) — technical document ation for

− pro viders of general-pur pose AI models

− Section 1

− Inf or mation to be provided by all providers of general-pur pose AI models

− The te chnical documentation refer red to in Ar ticle 53(1), point (a) shall contain at least the f ollowing inf or mation as

− appropr iate t o the size and r isk prof ile of the model:

− 1. A g eneral descr iption of the general-pur pose AI model including:

− (a) the tasks that the model is inte nded to perf or m and the type and nature of AI syste ms in which it can be

− integrat ed;

− (b) the accep table use policies applicable;

− (c) the dat e of release and methods of distr ibution;

− (d) the arch itec ture and number of parameter s;

− (e) the modality (e.g. text, imag e) and f or mat of in puts and outputs;

− (f) the licence.

− 2. A detailed descr iption of the elements of the model refe r red to in point 1, and relevant inf or mation of the process

− f or the development, including the f ollowi ng elements:

− (a) the t echnical means (e.g. instr uctions of use, infrastr ucture, tools) required f or the g eneral-pur pose AI model to

− be integrat ed in AI systems;

− (b) the design specific ations of the model and training process, including training methodologies and t echniques,

− the ke y design choices including the rationale and assum ptions made; what the model is designed to op timise f or

− and the relevance of the diffe rent parameters, as applicable;

− (c) inf or mation on the data used f or training, testing and validation, where applicable, including the type and

− prove nance of data and curation methodologies (e.g. cleaning, filter ing, etc.), the number of data points, their

− scope and main charact er istics; ho w the data was obtained and selecte d as well as all other measures to detect the

− unsuitability of data sources and methods to det ect identifia ble biases, where applicable;

− (d) the computational resources used to train the model (e.g. number of f loating point operations), training time,

− and other relevant details related to the training;

− (e) kno wn or estimated energy consump tion of the model.

− With regard to point (e), where the energy consump tion of the model is unknow n, the energy consump tion ma y be

− based on inf or mation about comp utational resources used.

− Section 2

− Ad ditional inf or mation to be provided b y provid ers of g eneral-pur pose AI models with syste mic r isk

− 1. A detailed descr ipt ion of the evaluation strategies, including evaluation results, on the basis of a vailable public

− evaluation prot ocols and t ools or other wise of other evaluation methodologies. Evaluation strategies shall include

− evaluation cr ite r ia, metr ics and the methodology on the identifica tion of limitations.

− 2. Where applicable, a detailed descr iption of the measures put in place f or the pur pose of conducting inter nal and/or

− external adversar ial te sting (e.g. red te aming), model adap tations, including alignment and fine -tuning.

− OJ L, 12.7.2024

− EN

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 141/144

− 3. Where applicable, a detailed descr iption of the system arch itecture explaining ho w sof tware com ponents build or

− f eed into each other and integrat e into the overa ll processing.

− EN

− OJ L, 12.7.2024

− 142/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

− ANNEX XII

− T ransparency infor mation refe r red t o in Ar ticle 53(1), point (b) — technical document ation for

− pro viders of general-pur pose AI models t o do wnstre am pro viders that integ rate the model into their

− AI system

− The inf or mation refer red to in Ar ticle 53(1), point (b) shall contain at least the f ollowing:

− 1. A g eneral descr iption of the general-pur pose AI model including:

− (a) the tasks that the model is inte nded to perfor m and the type and nature of AI systems into which it can be

− integrat ed;

− (b) the accep table use policies applicable;

− (c) the dat e of release and methods of distr ibution;

− (d) ho w the model interac ts, or can be used to interact, with hardware or sof tware that is not par t of the model

− itself, where applicable;

− (e) the versions of relevant sof tware related to the use of the general-pur pose AI model, where applicable;

− (f) the arch itec ture and number of parameter s;

− (g) the modality (e.g. text, imag e) and f or mat of in puts and outputs;

− (h) the licence f or the model.

− 2. A descr iption of the elements of the model and of the process f or its development, including:

− (a) the te ch nical means (e.g. instr uctions f or use, infrastr ucture, tools) required f or the general-pur pose AI model to

− be integrat ed into AI syste ms;

− (b) the modality (e.g. te xt, image, etc.) and f or mat of the inputs and outputs and their maximum size (e.g. cont ext

− window length, etc.);

− (c) inf or mation on the data used f or training, testing and validation, where applicable, including the type and

− prove nance of data and curation methodologies.

− OJ L, 12.7.2024

− EN

− ELI: http://data.europa.eu/eli/reg/2024/1689/oj 143/144

− Cr iter ia f or the designation of general-pur pose AI models with systemic r isk refe r red to in Ar ticle 51

− For the pur pose of det er mining that a g eneral-pur pose AI model has capabilities or an imp act equivalent to those set out in

− Ar ticle 51(1), point (a), the Commission shall take into account the f ollowing cr ite r ia:

− (a) the number of parameter s of the model;

− (b) the quality or size of the data set, f or example measured through to kens;

− (c) the amount of comp utation used f or training the model, measured in f loating point operations or indicate d by

− a combination of other var iables such as estimated cost of training, estimat ed time required f or the training, or

− estimated energy consump tion f or the training;

− (d) the input and output modalities of the model, suc h as text to te xt (large languag e models), te xt to imag e,

− multi-modality , and the state of the ar t thresholds f or determi ning high-impact capabilities f or each modality , and

− the specific type of in puts and outputs (e.g. biological sequences);

− (e) the benchmarks and evaluations of capabilities of the model, including consider ing the number of tasks without

− additional training, adap tability to lear n new , distinct tasks, its level of autonom y and scalability , the t ools it has

− access to ;

− (f) whether it has a high impact on the inte r nal marke t due to its reac h, which shall be presumed when it has been

− made ava ilable to at least 10 000 registered business users established in the Union;

− (g) the number of register ed end-users.

− EN

− OJ L, 12.7.2024

− 144/144 ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ Official Journal of the European Union

+ 2024/1689

+ REGULATION (EU) 2024/1689 OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL

+ EN L series

+ 12.7.2024

+ laying down harmonised rules on artificial intelligence and amending Regulations (EC) No 300/2008, (EU) No 167/2013, (EU) No 168/2013, (EU) 2018/858, (EU) 2018/1139 and (EU) 2019/2144 and Directives 2014/90/EU, (EU) 2016/797 and (EU) 2020/1828 (Artificial Intelligence Act)

+ (Text with EEA relevance)

+ THE EUROPEAN PARLIAMENT AND THE COUNCIL OF THE EUROPEAN UNION,

+ Having regard to the Treaty on the Functioning of the European Union, and in particular Articles 16 and 114 thereof,

+ After transmission of the draft legislative act to the national parliaments,

+ Having regard to the opinion of the European Economic and Social Committee (1),

+ Having regard to the opinion of the European Central Bank (2),

+ Having regard to the opinion of the Committee of the Regions (3),

+ Acting in accordance with the ordinary legislative procedure (4),

+ (1) The purpose of this Regulation is to improve the functioning of the internal market by laying down a uniform legal framework in particular for the development, the placing on the market, the putting into service and the use of artificial intelligence systems (AI systems) in the Union, in accordance with Union values, to promote the uptake of human centric and trustworthy artificial intelligence (AI) while ensuring a high level of protection of health, safety, fundamental rights as enshrined in the Charter of Fundamental Rights of the European Union (the `Charter'), including democracy, the rule of law and environmental protection, to protect against the harmful effects of AI systems in the Union, and to support innovation. This Regulation ensures the free movement, cross-border, of AI-based goods and services, thus preventing Member States from imposing restrictions on the development, marketing and use of AI systems, unless explicitly authorised by this Regulation.

+ (2) This Regulation should be applied in accordance with the values of the Union enshrined as in the Charter, facilitating the protection of natural persons, undertakings, democracy, the rule of law and environmental protection, while boosting innovation and employment and making the Union a leader in the uptake of trustworthy AI.

+ (3) AI systems can be easily deployed in a large variety of sectors of the economy and many parts of society, including across borders, and can easily circulate throughout the Union. Certain Member States have already explored the adoption of national rules to ensure that AI is trustworthy and safe and is developed and used in accordance with fundamental rights obligations. Diverging national rules may lead to the fragmentation of the internal market and may decrease legal certainty for operators that develop, import or use AI systems. A consistent and high level of protection throughout the Union should therefore be ensured in order to achieve trustworthy AI, while divergences hampering the free circulation, innovation, deployment and the uptake of AI systems and related products and services within the internal market should be prevented by laying down uniform obligations for operators and

+ (1) OJ C 517, 22.12.2021, p. 56. (2) OJ C 115, 11.3.2022, p. 5. (3) OJ C 97, 28.2.2022, p. 60. (4) Position of the European Parliament of 13 March 2024 (not yet published in the Official Journal) and decision of the Council of

+ 21 May 2024.

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 1/144

+ guaranteeing the uniform protection of overriding reasons of public interest and of rights of persons throughout the internal market on the basis of Article 114 of the Treaty on the Functioning of the European Union (TFEU). To the extent that this Regulation contains specific rules on the protection of individuals with regard to the processing of personal data concerning restrictions of the use of AI systems for remote biometric identification for the purpose of law enforcement, of the use of AI systems for risk assessments of natural persons for the purpose of law enforcement and of the use of AI systems of biometric categorisation for the purpose of law enforcement, it is appropriate to base this Regulation, in so far as those specific rules are concerned, on Article 16 TFEU. In light of those specific rules and the recourse to Article 16 TFEU, it is appropriate to consult the European Data Protection Board.

+ (4) AI is a fast evolving family of technologies that contributes to a wide array of economic, environmental and societal benefits across the entire spectrum of industries and social activities. By improving prediction, optimising operations and resource allocation, and personalising digital solutions available for individuals and organisations, the use of AI can provide key competitive advantages to undertakings and support socially and environmentally beneficial outcomes, for example in healthcare, agriculture, food safety, education and training, media, sports, culture, infrastructure management, energy, transport and logistics, public services, security, justice, resource and energy efficiency, environmental monitoring, the conservation and restoration of biodiversity and ecosystems and climate change mitigation and adaptation.

+ (5) At the same time, depending on the circumstances regarding its specific application, use, and level of technological development, AI may generate risks and cause harm to public interests and fundamental rights that are protected by Union law. Such harm might be material or immaterial, including physical, psychological, societal or economic harm.

+ (6) Given the major impact that AI can have on society and the need to build trust, it is vital for AI and its regulatory framework to be developed in accordance with Union values as enshrined in Article 2 of the Treaty on European Union (TEU), the fundamental rights and freedoms enshrined in the Treaties and, pursuant to Article 6 TEU, the Charter. As a prerequisite, AI should be a human-centric technology. It should serve as a tool for people, with the ultimate aim of increasing human well-being.

+ (7) In order to ensure a consistent and high level of protection of public interests as regards health, safety and fundamental rights, common rules for high-risk AI systems should be established. Those rules should be consistent with the Charter, non-discriminatory and in line with the Union's international trade commitments. They should also take into account the European Declaration on Digital Rights and Principles for the Digital Decade and the Ethics guidelines for trustworthy AI of the High-Level Expert Group on Artificial Intelligence (AI HLEG).

+ (8) A Union legal framework laying down harmonised rules on AI is therefore needed to foster the development, use and uptake of AI in the internal market that at the same time meets a high level of protection of public interests, such as health and safety and the protection of fundamental rights, including democracy, the rule of law and environmental protection as recognised and protected by Union law. To achieve that objective, rules regulating the placing on the market, the putting into service and the use of certain AI systems should be laid down, thus ensuring the smooth functioning of the internal market and allowing those systems to benefit from the principle of free movement of goods and services. Those rules should be clear and robust in protecting fundamental rights, supportive of new innovative solutions, enabling a European ecosystem of public and private actors creating AI systems in line with Union values and unlocking the potential of the digital transformation across all regions of the Union. By laying down those rules as well as measures in support of innovation with a particular focus on small and medium enterprises (SMEs), including startups, this Regulation supports the objective of promoting the European human-centric approach to AI and being a global leader in the development of secure, trustworthy and ethical AI as stated by the European Council (5), and it ensures the protection of ethical principles, as specifically requested by the European Parliament (6).

+ (5) European Council, Special meeting of the European Council (1 and 2 October 2020) -- Conclusions, EUCO 13/20, 2020, p. 6. (6) European Parliament resolution of 20 October 2020 with recommendations to the Commission on a framework of ethical aspects

+ of artificial intelligence, robotics and related technologies, 2020/2012(INL).

+ 2/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ (9) Harmonised rules applicable to the placing on the market, the putting into service and the use of high-risk AI systems should be laid down consistently with Regulation (EC) No 765/2008 of the European Parliament and of the Council (7), Decision No 768/2008/EC of the European Parliament and of the Council (8) and Regulation (EU) 2019/1020 of the European Parliament and of the Council (9) (New Legislative Framework). The harmonised rules laid down in this Regulation should apply across sectors and, in line with the New Legislative Framework, should be without prejudice to existing Union law, in particular on data protection, consumer protection, fundamental rights, employment, and protection of workers, and product safety, to which this Regulation is complementary. As a consequence, all rights and remedies provided for by such Union law to consumers, and other persons on whom AI systems may have a negative impact, including as regards the compensation of possible damages pursuant to Council Directive 85/374/EEC (10) remain unaffected and fully applicable. Furthermore, in the context of employment and protection of workers, this Regulation should therefore not affect Union law on social policy and national labour law, in compliance with Union law, concerning employment and working conditions, including health and safety at work and the relationship between employers and workers. This Regulation should also not affect the exercise of fundamental rights as recognised in the Member States and at Union level, including the right or freedom to strike or to take other action covered by the specific industrial relations systems in Member States as well as the right to negotiate, to conclude and enforce collective agreements or to take collective action in accordance with national law. This Regulation should not affect the provisions aiming to improve working conditions in platform work laid down in a Directive of the European Parliament and of the Council on improving working conditions in platform work. Moreover, this Regulation aims to strengthen the effectiveness of such existing rights and remedies by establishing specific requirements and obligations, including in respect of the transparency, technical documentation and record-keeping of AI systems. Furthermore, the obligations placed on various operators involved in the AI value chain under this Regulation should apply without prejudice to national law, in compliance with Union law, having the effect of limiting the use of certain AI systems where such law falls outside the scope of this Regulation or pursues legitimate public interest objectives other than those pursued by this Regulation. For example, national labour law and law on the protection of minors, namely persons below the age of 18, taking into account the UNCRC General Comment No 25 (2021) on children's rights in relation to the digital environment, insofar as they are not specific to AI systems and pursue other legitimate public interest objectives, should not be affected by this Regulation.

+ (10) The fundamental right to the protection of personal data is safeguarded in particular by Regulations (EU) 2016/679 (11) and (EU) 2018/1725 (12) of the European Parliament and of the Council and Directive (EU) 2016/680 of the European Parliament and of the Council (13). Directive 2002/58/EC of the European Parliament and of the Council (14) additionally protects private life and the confidentiality of communications, including by way of providing conditions for any storing of personal and non-personal data in, and access from, terminal equipment. Those Union legal acts provide the basis for sustainable and responsible data processing, including where data sets include a mix of personal and non-personal data. This Regulation does not seek to affect the application of existing Union law governing the processing of personal data, including the tasks and powers of the independent supervisory authorities competent to monitor compliance with those instruments. It also does not affect the obligations of providers and deployers of AI systems in their role as data controllers or processors stemming from Union or national law on the protection of personal data in so far as the design, the development or the use of AI systems involves the processing of personal data. It is also appropriate to clarify that data subjects continue to enjoy all the

+ (7) Regulation (EC) No 765/2008 of the European Parliament and of the Council of 9 July 2008 setting out the requirements for accreditation and repealing Regulation (EEC) No 339/93 (OJ L 218, 13.8.2008, p. 30).

+ (8) Decision No 768/2008/EC of the European Parliament and of the Council of 9 July 2008 on a common framework for the marketing of products, and repealing Council Decision 93/465/EEC (OJ L 218, 13.8.2008, p. 82).

+ (9) Regulation (EU) 2019/1020 of the European Parliament and of the Council of 20 June 2019 on market surveillance and compliance of products and amending Directive 2004/42/EC and Regulations (EC) No 765/2008 and (EU) No 305/2011 (OJ L 169, 25.6.2019, p. 1).

+ (10) Council Directive 85/374/EEC of 25 July 1985 on the approximation of the laws, regulations and administrative provisions of the Member States concerning liability for defective products (OJ L 210, 7.8.1985, p. 29).

+ (11) Regulation (EU) 2016/679 of the European Parliament and of the Council of 27 April 2016 on the protection of natural persons with regard to the processing of personal data and on the free movement of such data, and repealing Directive 95/46/EC (General Data Protection Regulation) (OJ L 119, 4.5.2016, p. 1).

+ (12) Regulation (EU) 2018/1725 of the European Parliament and of the Council of 23 October 2018 on the protection of natural persons with regard to the processing of personal data by the Union institutions, bodies, offices and agencies and on the free movement of such data, and repealing Regulation (EC) No 45/2001 and Decision No 1247/2002/EC (OJ L 295, 21.11.2018, p. 39).

+ (13) Directive (EU) 2016/680 of the European Parliament and of the Council of 27 April 2016 on the protection of natural persons with regard to the processing of personal data by competent authorities for the purposes of the prevention, investigation, detection or prosecution of criminal offences or the execution of criminal penalties, and on the free movement of such data, and repealing Council Framework Decision 2008/977/JHA (OJ L 119, 4.5.2016, p. 89).

+ (14) Directive 2002/58/EC of the European Parliament and of the Council of 12 July 2002 concerning the processing of personal data and the protection of privacy in the electronic communications sector (Directive on privacy and electronic communications) (OJ L 201, 31.7.2002, p. 37).

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 3/144

+ rights and guarantees awarded to them by such Union law, including the rights related to solely automated individual decision-making, including profiling. Harmonised rules for the placing on the market, the putting into service and the use of AI systems established under this Regulation should facilitate the effective implementation and enable the exercise of the data subjects' rights and other remedies guaranteed under Union law on the protection of personal data and of other fundamental rights.

+ (11) This Regulation should be without prejudice to the provisions regarding the liability of providers of intermediary services as set out in Regulation (EU) 2022/2065 of the European Parliament and of the Council (15).

+ (12) The notion of `AI system' in this Regulation should be clearly defined and should be closely aligned with the work of international organisations working on AI to ensure legal certainty, facilitate international convergence and wide acceptance, while providing the flexibility to accommodate the rapid technological developments in this field. Moreover, the definition should be based on key characteristics of AI systems that distinguish it from simpler traditional software systems or programming approaches and should not cover systems that are based on the rules defined solely by natural persons to automatically execute operations. A key characteristic of AI systems is their capability to infer. This capability to infer refers to the process of obtaining the outputs, such as predictions, content, recommendations, or decisions, which can influence physical and virtual environments, and to a capability of AI systems to derive models or algorithms, or both, from inputs or data. The techniques that enable inference while building an AI system include machine learning approaches that learn from data how to achieve certain objectives, and logic- and knowledge-based approaches that infer from encoded knowledge or symbolic representation of the task to be solved. The capacity of an AI system to infer transcends basic data processing by enabling learning, reasoning or modelling. The term `machine-based' refers to the fact that AI systems run on machines. The reference to explicit or implicit objectives underscores that AI systems can operate according to explicit defined objectives or to implicit objectives. The objectives of the AI system may be different from the intended purpose of the AI system in a specific context. For the purposes of this Regulation, environments should be understood to be the contexts in which the AI systems operate, whereas outputs generated by the AI system reflect different functions performed by AI systems and include predictions, content, recommendations or decisions. AI systems are designed to operate with varying levels of autonomy, meaning that they have some degree of independence of actions from human involvement and of capabilities to operate without human intervention. The adaptiveness that an AI system could exhibit after deployment, refers to self-learning capabilities, allowing the system to change while in use. AI systems can be used on a stand-alone basis or as a component of a product, irrespective of whether the system is physically integrated into the product (embedded) or serves the functionality of the product without being integrated therein (non-embedded).

+ (13) The notion of `deployer' referred to in this Regulation should be interpreted as any natural or legal person, including a public authority, agency or other body, using an AI system under its authority, except where the AI system is used in the course of a personal non-professional activity. Depending on the type of AI system, the use of the system may affect persons other than the deployer.

+ (14) The notion of `biometric data' used in this Regulation should be interpreted in light of the notion of biometric data as defined in Article 4, point (14) of Regulation (EU) 2016/679, Article 3, point (18) of Regulation (EU) 2018/1725 and Article 3, point (13) of Directive (EU) 2016/680. Biometric data can allow for the authentication, identification or categorisation of natural persons and for the recognition of emotions of natural persons.

+ (15) The notion of `biometric identification' referred to in this Regulation should be defined as the automated recognition of physical, physiological and behavioural human features such as the face, eye movement, body shape, voice, prosody, gait, posture, heart rate, blood pressure, odour, keystrokes characteristics, for the purpose of establishing an individual's identity by comparing biometric data of that individual to stored biometric data of individuals in a reference database, irrespective of whether the individual has given its consent or not. This excludes AI systems intended to be used for biometric verification, which includes authentication, whose sole purpose is to confirm that a specific natural person is the person he or she claims to be and to confirm the identity of a natural person for the sole purpose of having access to a service, unlocking a device or having security access to premises.

+ (15) Regulation (EU) 2022/2065 of the European Parliament and of the Council of 19 October 2022 on a Single Market For Digital Services and amending Directive 2000/31/EC (Digital Services Act) (OJ L 277, 27.10.2022, p. 1).

+ 4/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ (16) The notion of `biometric categorisation' referred to in this Regulation should be defined as assigning natural persons to specific categories on the basis of their biometric data. Such specific categories can relate to aspects such as sex, age, hair colour, eye colour, tattoos, behavioural or personality traits, language, religion, membership of a national minority, sexual or political orientation. This does not include biometric categorisation systems that are a purely ancillary feature intrinsically linked to another commercial service, meaning that the feature cannot, for objective technical reasons, be used without the principal service, and the integration of that feature or functionality is not a means to circumvent the applicability of the rules of this Regulation. For example, filters categorising facial or body features used on online marketplaces could constitute such an ancillary feature as they can be used only in relation to the principal service which consists in selling a product by allowing the consumer to preview the display of the product on him or herself and help the consumer to make a purchase decision. Filters used on online social network services which categorise facial or body features to allow users to add or modify pictures or videos could also be considered to be ancillary feature as such filter cannot be used without the principal service of the social network services consisting in the sharing of content online.

+ (17) The notion of `remote biometric identification system' referred to in this Regulation should be defined functionally, as an AI system intended for the identification of natural persons without their active involvement, typically at a distance, through the comparison of a person's biometric data with the biometric data contained in a reference database, irrespectively of the particular technology, processes or types of biometric data used. Such remote biometric identification systems are typically used to perceive multiple persons or their behaviour simultaneously in order to facilitate significantly the identification of natural persons without their active involvement. This excludes AI systems intended to be used for biometric verification, which includes authentication, the sole purpose of which is to confirm that a specific natural person is the person he or she claims to be and to confirm the identity of a natural person for the sole purpose of having access to a service, unlocking a device or having security access to premises. That exclusion is justified by the fact that such systems are likely to have a minor impact on fundamental rights of natural persons compared to the remote biometric identification systems which may be used for the processing of the biometric data of a large number of persons without their active involvement. In the case of `real-time' systems, the capturing of the biometric data, the comparison and the identification occur all instantaneously, near-instantaneously or in any event without a significant delay. In this regard, there should be no scope for circumventing the rules of this Regulation on the `real-time' use of the AI systems concerned by providing for minor delays. `Real-time' systems involve the use of `live' or `near-live' material, such as video footage, generated by a camera or other device with similar functionality. In the case of `post' systems, in contrast, the biometric data has already been captured and the comparison and identification occur only after a significant delay. This involves material, such as pictures or video footage generated by closed circuit television cameras or private devices, which has been generated before the use of the system in respect of the natural persons concerned.

+ (18) The notion of `emotion recognition system' referred to in this Regulation should be defined as an AI system for the purpose of identifying or inferring emotions or intentions of natural persons on the basis of their biometric data. The notion refers to emotions or intentions such as happiness, sadness, anger, surprise, disgust, embarrassment, excitement, shame, contempt, satisfaction and amusement. It does not include physical states, such as pain or fatigue, including, for example, systems used in detecting the state of fatigue of professional pilots or drivers for the purpose of preventing accidents. This does also not include the mere detection of readily apparent expressions, gestures or movements, unless they are used for identifying or inferring emotions. Those expressions can be basic facial expressions, such as a frown or a smile, or gestures such as the movement of hands, arms or head, or characteristics of a person's voice, such as a raised voice or whispering.

+ (19) For the purposes of this Regulation the notion of `publicly accessible space' should be understood as referring to any physical space that is accessible to an undetermined number of natural persons, and irrespective of whether the space in question is privately or publicly owned, irrespective of the activity for which the space may be used, such as for commerce, for example, shops, restaurants, caf�s; for services, for example, banks, professional activities, hospitality; for sport, for example, swimming pools, gyms, stadiums; for transport, for example, bus, metro and railway stations, airports, means of transport; for entertainment, for example, cinemas, theatres, museums, concert and conference halls; or for leisure or otherwise, for example, public roads and squares, parks, forests, playgrounds. A space should also be classified as being publicly accessible if, regardless of potential capacity or security restrictions, access is subject to certain predetermined conditions which can be fulfilled by an undetermined number of persons, such as the purchase of a ticket or title of transport, prior registration or having a certain age. In contrast, a space should not be considered to be publicly accessible if access is limited to specific and defined natural persons through either Union or national law directly related to public safety or security or through the clear manifestation

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 5/144

+ of will by the person having the relevant authority over the space. The factual possibility of access alone, such as an unlocked door or an open gate in a fence, does not imply that the space is publicly accessible in the presence of indications or circumstances suggesting the contrary, such as. signs prohibiting or restricting access. Company and factory premises, as well as offices and workplaces that are intended to be accessed only by relevant employees and service providers, are spaces that are not publicly accessible. Publicly accessible spaces should not include prisons or border control. Some other spaces may comprise both publicly accessible and non-publicly accessible spaces, such as the hallway of a private residential building necessary to access a doctor's office or an airport. Online spaces are not covered, as they are not physical spaces. Whether a given space is accessible to the public should however be determined on a case-by-case basis, having regard to the specificities of the individual situation at hand.

+ (20) In order to obtain the greatest benefits from AI systems while protecting fundamental rights, health and safety and to enable democratic control, AI literacy should equip providers, deployers and affected persons with the necessary notions to make informed decisions regarding AI systems. Those notions may vary with regard to the relevant context and can include understanding the correct application of technical elements during the AI system's development phase, the measures to be applied during its use, the suitable ways in which to interpret the AI system's output, and, in the case of affected persons, the knowledge necessary to understand how decisions taken with the assistance of AI will have an impact on them. In the context of the application this Regulation, AI literacy should provide all relevant actors in the AI value chain with the insights required to ensure the appropriate compliance and its correct enforcement. Furthermore, the wide implementation of AI literacy measures and the introduction of appropriate follow-up actions could contribute to improving working conditions and ultimately sustain the consolidation, and innovation path of trustworthy AI in the Union. The European Artificial Intelligence Board (the `Board') should support the Commission, to promote AI literacy tools, public awareness and understanding of the benefits, risks, safeguards, rights and obligations in relation to the use of AI systems. In cooperation with the relevant stakeholders, the Commission and the Member States should facilitate the drawing up of voluntary codes of conduct to advance AI literacy among persons dealing with the development, operation and use of AI.

+ (21) In order to ensure a level playing field and an effective protection of rights and freedoms of individuals across the Union, the rules established by this Regulation should apply to providers of AI systems in a non-discriminatory manner, irrespective of whether they are established within the Union or in a third country, and to deployers of AI systems established within the Union.

+ (22) In light of their digital nature, certain AI systems should fall within the scope of this Regulation even when they are not placed on the market, put into service, or used in the Union. This is the case, for example, where an operator established in the Union contracts certain services to an operator established in a third country in relation to an activity to be performed by an AI system that would qualify as high-risk. In those circumstances, the AI system used in a third country by the operator could process data lawfully collected in and transferred from the Union, and provide to the contracting operator in the Union the output of that AI system resulting from that processing, without that AI system being placed on the market, put into service or used in the Union. To prevent the circumvention of this Regulation and to ensure an effective protection of natural persons located in the Union, this Regulation should also apply to providers and deployers of AI systems that are established in a third country, to the extent the output produced by those systems is intended to be used in the Union. Nonetheless, to take into account existing arrangements and special needs for future cooperation with foreign partners with whom information and evidence is exchanged, this Regulation should not apply to public authorities of a third country and international organisations when acting in the framework of cooperation or international agreements concluded at Union or national level for law enforcement and judicial cooperation with the Union or the Member States, provided that the relevant third country or international organisation provides adequate safeguards with respect to the protection of fundamental rights and freedoms of individuals. Where relevant, this may cover activities of entities entrusted by the third countries to carry out specific tasks in support of such law enforcement and judicial cooperation. Such framework for cooperation or agreements have been established bilaterally between Member States and third countries or between the European Union, Europol and other Union agencies and third countries and international organisations. The authorities competent for supervision of the law enforcement and judicial authorities under this Regulation should assess whether those frameworks for cooperation or international agreements include adequate safeguards with respect to the protection of fundamental rights and freedoms of individuals. Recipient national

+ 6/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ authorities and Union institutions, bodies, offices and agencies making use of such outputs in the Union remain accountable to ensure their use complies with Union law. When those international agreements are revised or new ones are concluded in the future, the contracting parties should make utmost efforts to align those agreements with the requirements of this Regulation.

+ (23) This Regulation should also apply to Union institutions, bodies, offices and agencies when acting as a provider or deployer of an AI system.

+ (24) If, and insofar as, AI systems are placed on the market, put into service, or used with or without modification of such systems for military, defence or national security purposes, those should be excluded from the scope of this Regulation regardless of which type of entity is carrying out those activities, such as whether it is a public or private entity. As regards military and defence purposes, such exclusion is justified both by Article 4(2) TEU and by the specificities of the Member States' and the common Union defence policy covered by Chapter 2 of Title V TEU that are subject to public international law, which is therefore the more appropriate legal framework for the regulation of AI systems in the context of the use of lethal force and other AI systems in the context of military and defence activities. As regards national security purposes, the exclusion is justified both by the fact that national security remains the sole responsibility of Member States in accordance with Article 4(2) TEU and by the specific nature and operational needs of national security activities and specific national rules applicable to those activities. Nonetheless, if an AI system developed, placed on the market, put into service or used for military, defence or national security purposes is used outside those temporarily or permanently for other purposes, for example, civilian or humanitarian purposes, law enforcement or public security purposes, such a system would fall within the scope of this Regulation. In that case, the entity using the AI system for other than military, defence or national security purposes should ensure the compliance of the AI system with this Regulation, unless the system is already compliant with this Regulation. AI systems placed on the market or put into service for an excluded purpose, namely military, defence or national security, and one or more non-excluded purposes, such as civilian purposes or law enforcement, fall within the scope of this Regulation and providers of those systems should ensure compliance with this Regulation. In those cases, the fact that an AI system may fall within the scope of this Regulation should not affect the possibility of entities carrying out national security, defence and military activities, regardless of the type of entity carrying out those activities, to use AI systems for national security, military and defence purposes, the use of which is excluded from the scope of this Regulation. An AI system placed on the market for civilian or law enforcement purposes which is used with or without modification for military, defence or national security purposes should not fall within the scope of this Regulation, regardless of the type of entity carrying out those activities.

+ (25) This Regulation should support innovation, should respect freedom of science, and should not undermine research and development activity. It is therefore necessary to exclude from its scope AI systems and models specifically developed and put into service for the sole purpose of scientific research and development. Moreover, it is necessary to ensure that this Regulation does not otherwise affect scientific research and development activity on AI systems or models prior to being placed on the market or put into service. As regards product-oriented research, testing and development activity regarding AI systems or models, the provisions of this Regulation should also not apply prior to those systems and models being put into service or placed on the market. That exclusion is without prejudice to the obligation to comply with this Regulation where an AI system falling into the scope of this Regulation is placed on the market or put into service as a result of such research and development activity and to the application of provisions on AI regulatory sandboxes and testing in real world conditions. Furthermore, without prejudice to the exclusion of AI systems specifically developed and put into service for the sole purpose of scientific research and development, any other AI system that may be used for the conduct of any research and development activity should remain subject to the provisions of this Regulation. In any event, any research and development activity should be carried out in accordance with recognised ethical and professional standards for scientific research and should be conducted in accordance with applicable Union law.

+ (26) In order to introduce a proportionate and effective set of binding rules for AI systems, a clearly defined risk-based approach should be followed. That approach should tailor the type and content of such rules to the intensity and scope of the risks that AI systems can generate. It is therefore necessary to prohibit certain unacceptable AI practices, to lay down requirements for high-risk AI systems and obligations for the relevant operators, and to lay down transparency obligations for certain AI systems.

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 7/144

+ (27) While the risk-based approach is the basis for a proportionate and effective set of binding rules, it is important to recall the 2019 Ethics guidelines for trustworthy AI developed by the independent AI HLEG appointed by the Commission. In those guidelines, the AI HLEG developed seven non-binding ethical principles for AI which are intended to help ensure that AI is trustworthy and ethically sound. The seven principles include human agency and oversight; technical robustness and safety; privacy and data governance; transparency; diversity, non-discrimination and fairness; societal and environmental well-being and accountability. Without prejudice to the legally binding requirements of this Regulation and any other applicable Union law, those guidelines contribute to the design of coherent, trustworthy and human-centric AI, in line with the Charter and with the values on which the Union is founded. According to the guidelines of the AI HLEG, human agency and oversight means that AI systems are developed and used as a tool that serves people, respects human dignity and personal autonomy, and that is functioning in a way that can be appropriately controlled and overseen by humans. Technical robustness and safety means that AI systems are developed and used in a way that allows robustness in the case of problems and resilience against attempts to alter the use or performance of the AI system so as to allow unlawful use by third parties, and minimise unintended harm. Privacy and data governance means that AI systems are developed and used in accordance with privacy and data protection rules, while processing data that meets high standards in terms of quality and integrity. Transparency means that AI systems are developed and used in a way that allows appropriate traceability and explainability, while making humans aware that they communicate or interact with an AI system, as well as duly informing deployers of the capabilities and limitations of that AI system and affected persons about their rights. Diversity, non-discrimination and fairness means that AI systems are developed and used in a way that includes diverse actors and promotes equal access, gender equality and cultural diversity, while avoiding discriminatory impacts and unfair biases that are prohibited by Union or national law. Social and environmental well-being means that AI systems are developed and used in a sustainable and environmentally friendly manner as well as in a way to benefit all human beings, while monitoring and assessing the long-term impacts on the individual, society and democracy. The application of those principles should be translated, when possible, in the design and use of AI models. They should in any case serve as a basis for the drafting of codes of conduct under this Regulation. All stakeholders, including industry, academia, civil society and standardisation organisations, are encouraged to take into account, as appropriate, the ethical principles for the development of voluntary best practices and standards.

+ (28) Aside from the many beneficial uses of AI, it can also be misused and provide novel and powerful tools for manipulative, exploitative and social control practices. Such practices are particularly harmful and abusive and should be prohibited because they contradict Union values of respect for human dignity, freedom, equality, democracy and the rule of law and fundamental rights enshrined in the Charter, including the right to non-discrimination, to data protection and to privacy and the rights of the child.

+ (29) AI-enabled manipulative techniques can be used to persuade persons to engage in unwanted behaviours, or to deceive them by nudging them into decisions in a way that subverts and impairs their autonomy, decision-making and free choices. The placing on the market, the putting into service or the use of certain AI systems with the objective to or the effect of materially distorting human behaviour, whereby significant harms, in particular having sufficiently important adverse impacts on physical, psychological health or financial interests are likely to occur, are particularly dangerous and should therefore be prohibited. Such AI systems deploy subliminal components such as audio, image, video stimuli that persons cannot perceive, as those stimuli are beyond human perception, or other manipulative or deceptive techniques that subvert or impair person's autonomy, decision-making or free choice in ways that people are not consciously aware of those techniques or, where they are aware of them, can still be deceived or are not able to control or resist them. This could be facilitated, for example, by machine-brain interfaces or virtual reality as they allow for a higher degree of control of what stimuli are presented to persons, insofar as they may materially distort their behaviour in a significantly harmful manner. In addition, AI systems may also otherwise exploit the vulnerabilities of a person or a specific group of persons due to their age, disability within the meaning of Directive (EU) 2019/882 of the European Parliament and of the Council (16), or a specific social or economic situation that is likely to make those persons more vulnerable to exploitation such as persons living in extreme poverty, ethnic or religious minorities. Such AI systems can be placed on the market, put into service or used with the objective to or the effect of materially distorting the behaviour of a person and in a manner that causes or is reasonably likely to cause significant harm to that or another person or groups of persons, including harms that may be accumulated over time and should therefore be prohibited. It may not be possible to assume that there is an

+ (16) Directive (EU) 2019/882 of the European Parliament and of the Council of 17 April 2019 on the accessibility requirements for products and services (OJ L 151, 7.6.2019, p. 70).

+ 8/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ intention to distort behaviour where the distortion results from factors external to the AI system which are outside the control of the provider or the deployer, namely factors that may not be reasonably foreseeable and therefore not possible for the provider or the deployer of the AI system to mitigate. In any case, it is not necessary for the provider or the deployer to have the intention to cause significant harm, provided that such harm results from the manipulative or exploitative AI-enabled practices. The prohibitions for such AI practices are complementary to the provisions contained in Directive 2005/29/EC of the European Parliament and of the Council (17), in particular unfair commercial practices leading to economic or financial harms to consumers are prohibited under all circumstances, irrespective of whether they are put in place through AI systems or otherwise. The prohibitions of manipulative and exploitative practices in this Regulation should not affect lawful practices in the context of medical treatment such as psychological treatment of a mental disease or physical rehabilitation, when those practices are carried out in accordance with the applicable law and medical standards, for example explicit consent of the individuals or their legal representatives. In addition, common and legitimate commercial practices, for example in the field of advertising, that comply with the applicable law should not, in themselves, be regarded as constituting harmful manipulative AI-enabled practices.

+ (30) Biometric categorisation systems that are based on natural persons' biometric data, such as an individual person's face or fingerprint, to deduce or infer an individuals' political opinions, trade union membership, religious or philosophical beliefs, race, sex life or sexual orientation should be prohibited. That prohibition should not cover the lawful labelling, filtering or categorisation of biometric data sets acquired in line with Union or national law according to biometric data, such as the sorting of images according to hair colour or eye colour, which can for example be used in the area of law enforcement.

+ (31) AI systems providing social scoring of natural persons by public or private actors may lead to discriminatory outcomes and the exclusion of certain groups. They may violate the right to dignity and non-discrimination and the values of equality and justice. Such AI systems evaluate or classify natural persons or groups thereof on the basis of multiple data points related to their social behaviour in multiple contexts or known, inferred or predicted personal or personality characteristics over certain periods of time. The social score obtained from such AI systems may lead to the detrimental or unfavourable treatment of natural persons or whole groups thereof in social contexts, which are unrelated to the context in which the data was originally generated or collected or to a detrimental treatment that is disproportionate or unjustified to the gravity of their social behaviour. AI systems entailing such unacceptable scoring practices and leading to such detrimental or unfavourable outcomes should therefore be prohibited. That prohibition should not affect lawful evaluation practices of natural persons that are carried out for a specific purpose in accordance with Union and national law.

+ (32) The use of AI systems for `real-time' remote biometric identification of natural persons in publicly accessible spaces for the purpose of law enforcement is particularly intrusive to the rights and freedoms of the concerned persons, to the extent that it may affect the private life of a large part of the population, evoke a feeling of constant surveillance and indirectly dissuade the exercise of the freedom of assembly and other fundamental rights. Technical inaccuracies of AI systems intended for the remote biometric identification of natural persons can lead to biased results and entail discriminatory effects. Such possible biased results and discriminatory effects are particularly relevant with regard to age, ethnicity, race, sex or disabilities. In addition, the immediacy of the impact and the limited opportunities for further checks or corrections in relation to the use of such systems operating in real-time carry heightened risks for the rights and freedoms of the persons concerned in the context of, or impacted by, law enforcement activities.

+ (33) The use of those systems for the purpose of law enforcement should therefore be prohibited, except in exhaustively listed and narrowly defined situations, where the use is strictly necessary to achieve a substantial public interest, the importance of which outweighs the risks. Those situations involve the search for certain victims of crime including missing persons; certain threats to the life or to the physical safety of natural persons or of a terrorist attack; and the localisation or identification of perpetrators or suspects of the criminal offences listed in an annex to this Regulation, where those criminal offences are punishable in the Member State concerned by a custodial sentence or a detention

+ (17) Directive 2005/29/EC of the European Parliament and of the Council of 11 May 2005 concerning unfair business-to-consumer commercial practices in the internal market and amending Council Directive 84/450/EEC, Directives 97/7/EC, 98/27/EC and 2002/65/EC of the European Parliament and of the Council and Regulation (EC) No 2006/2004 of the European Parliament and of the Council (`Unfair Commercial Practices Directive') (OJ L 149, 11.6.2005, p. 22).

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 9/144

+ order for a maximum period of at least four years and as they are defined in the law of that Member State. Such a threshold for the custodial sentence or detention order in accordance with national law contributes to ensuring that the offence should be serious enough to potentially justify the use of `real-time' remote biometric identification systems. Moreover, the list of criminal offences provided in an annex to this Regulation is based on the 32 criminal offences listed in the Council Framework Decision 2002/584/JHA (18), taking into account that some of those offences are, in practice, likely to be more relevant than others, in that the recourse to `real-time' remote biometric identification could, foreseeably, be necessary and proportionate to highly varying degrees for the practical pursuit of the localisation or identification of a perpetrator or suspect of the different criminal offences listed and having regard to the likely differences in the seriousness, probability and scale of the harm or possible negative consequences. An imminent threat to life or the physical safety of natural persons could also result from a serious disruption of critical infrastructure, as defined in Article 2, point (4) of Directive (EU) 2022/2557 of the European Parliament and of the Council (19), where the disruption or destruction of such critical infrastructure would result in an imminent threat to life or the physical safety of a person, including through serious harm to the provision of basic supplies to the population or to the exercise of the core function of the State. In addition, this Regulation should preserve the ability for law enforcement, border control, immigration or asylum authorities to carry out identity checks in the presence of the person concerned in accordance with the conditions set out in Union and national law for such checks. In particular, law enforcement, border control, immigration or asylum authorities should be able to use information systems, in accordance with Union or national law, to identify persons who, during an identity check, either refuse to be identified or are unable to state or prove their identity, without being required by this Regulation to obtain prior authorisation. This could be, for example, a person involved in a crime, being unwilling, or unable due to an accident or a medical condition, to disclose their identity to law enforcement authorities.

+ (34) In order to ensure that those systems are used in a responsible and proportionate manner, it is also important to establish that, in each of those exhaustively listed and narrowly defined situations, certain elements should be taken into account, in particular as regards the nature of the situation giving rise to the request and the consequences of the use for the rights and freedoms of all persons concerned and the safeguards and conditions provided for with the use. In addition, the use of `real-time' remote biometric identification systems in publicly accessible spaces for the purpose of law enforcement should be deployed only to confirm the specifically targeted individual's identity and should be limited to what is strictly necessary concerning the period of time, as well as the geographic and personal scope, having regard in particular to the evidence or indications regarding the threats, the victims or perpetrator. The use of the real-time remote biometric identification system in publicly accessible spaces should be authorised only if the relevant law enforcement authority has completed a fundamental rights impact assessment and, unless provided otherwise in this Regulation, has registered the system in the database as set out in this Regulation. The reference database of persons should be appropriate for each use case in each of the situations mentioned above.

+ (35) Each use of a `real-time' remote biometric identification system in publicly accessible spaces for the purpose of law enforcement should be subject to an express and specific authorisation by a judicial authority or by an independent administrative authority of a Member State whose decision is binding. Such authorisation should, in principle, be obtained prior to the use of the AI system with a view to identifying a person or persons. Exceptions to that rule should be allowed in duly justified situations on grounds of urgency, namely in situations where the need to use the systems concerned is such as to make it effectively and objectively impossible to obtain an authorisation before commencing the use of the AI system. In such situations of urgency, the use of the AI system should be restricted to the absolute minimum necessary and should be subject to appropriate safeguards and conditions, as determined in national law and specified in the context of each individual urgent use case by the law enforcement authority itself. In addition, the law enforcement authority should in such situations request such authorisation while providing the reasons for not having been able to request it earlier, without undue delay and at the latest within 24 hours. If such an authorisation is rejected, the use of real-time biometric identification systems linked to that authorisation should cease with immediate effect and all the data related to such use should be discarded and deleted. Such data includes

+ (18) Council Framework Decision 2002/584/JHA of 13 June 2002 on the European arrest warrant and the surrender procedures between Member States (OJ L 190, 18.7.2002, p. 1).

+ (19) Directive (EU) 2022/2557 of the European Parliament and of the Council of 14 December 2022 on the resilience of critical entities and repealing Council Directive 2008/114/EC (OJ L 333, 27.12.2022, p. 164).

+ 10/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ input data directly acquired by an AI system in the course of the use of such system as well as the results and outputs of the use linked to that authorisation. It should not include input that is legally acquired in accordance with another Union or national law. In any case, no decision producing an adverse legal effect on a person should be taken based solely on the output of the remote biometric identification system.

+ (36) In order to carry out their tasks in accordance with the requirements set out in this Regulation as well as in national rules, the relevant market surveillance authority and the national data protection authority should be notified of each use of the real-time biometric identification system. Market surveillance authorities and the national data protection authorities that have been notified should submit to the Commission an annual report on the use of real-time biometric identification systems.

+ (37) Furthermore, it is appropriate to provide, within the exhaustive framework set by this Regulation that such use in the territory of a Member State in accordance with this Regulation should only be possible where and in as far as the Member State concerned has decided to expressly provide for the possibility to authorise such use in its detailed rules of national law. Consequently, Member States remain free under this Regulation not to provide for such a possibility at all or to only provide for such a possibility in respect of some of the objectives capable of justifying authorised use identified in this Regulation. Such national rules should be notified to the Commission within 30 days of their adoption.

+ (38) The use of AI systems for real-time remote biometric identification of natural persons in publicly accessible spaces for the purpose of law enforcement necessarily involves the processing of biometric data. The rules of this Regulation that prohibit, subject to certain exceptions, such use, which are based on Article 16 TFEU, should apply as lex specialis in respect of the rules on the processing of biometric data contained in Article 10 of Directive (EU) 2016/680, thus regulating such use and the processing of biometric data involved in an exhaustive manner. Therefore, such use and processing should be possible only in as far as it is compatible with the framework set by this Regulation, without there being scope, outside that framework, for the competent authorities, where they act for purpose of law enforcement, to use such systems and process such data in connection thereto on the grounds listed in Article 10 of Directive (EU) 2016/680. In that context, this Regulation is not intended to provide the legal basis for the processing of personal data under Article 8 of Directive (EU) 2016/680. However, the use of real-time remote biometric identification systems in publicly accessible spaces for purposes other than law enforcement, including by competent authorities, should not be covered by the specific framework regarding such use for the purpose of law enforcement set by this Regulation. Such use for purposes other than law enforcement should therefore not be subject to the requirement of an authorisation under this Regulation and the applicable detailed rules of national law that may give effect to that authorisation.

+ (39) Any processing of biometric data and other personal data involved in the use of AI systems for biometric identification, other than in connection to the use of real-time remote biometric identification systems in publicly accessible spaces for the purpose of law enforcement as regulated by this Regulation, should continue to comply with all requirements resulting from Article 10 of Directive (EU) 2016/680. For purposes other than law enforcement, Article 9(1) of Regulation (EU) 2016/679 and Article 10(1) of Regulation (EU) 2018/1725 prohibit the processing of biometric data subject to limited exceptions as provided in those Articles. In the application of Article 9(1) of Regulation (EU) 2016/679, the use of remote biometric identification for purposes other than law enforcement has already been subject to prohibition decisions by national data protection authorities.

+ (40) In accordance with Article 6a of Protocol No 21 on the position of the United Kingdom and Ireland in respect of the area of freedom, security and justice, as annexed to the TEU and to the TFEU, Ireland is not bound by the rules laid down in Article 5(1), first subparagraph, point (g), to the extent it applies to the use of biometric categorisation systems for activities in the field of police cooperation and judicial cooperation in criminal matters, Article 5(1), first subparagraph, point (d), to the extent it applies to the use of AI systems covered by that provision, Article 5(1), first subparagraph, point (h), Article 5(2) to (6) and Article 26(10) of this Regulation adopted on the basis of Article 16 TFEU which relate to the processing of personal data by the Member States when carrying out activities falling within the scope of Chapter 4 or Chapter 5 of Title V of Part Three of the TFEU, where Ireland is not bound by the rules governing the forms of judicial cooperation in criminal matters or police cooperation which require compliance with the provisions laid down on the basis of Article 16 TFEU.

+ (41) In accordance with Articles 2 and 2a of Protocol No 22 on the position of Denmark, annexed to the TEU and to the TFEU, Denmark is not bound by rules laid down in Article 5(1), first subparagraph, point (g), to the extent it applies to the use of biometric categorisation systems for activities in the field of police cooperation and judicial cooperation in criminal matters, Article 5(1), first subparagraph, point (d), to the extent it applies to the use of AI systems covered by that provision, Article 5(1), first subparagraph, point (h), (2) to (6) and Article 26(10) of this Regulation

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 11/144

+ adopted on the basis of Article 16 TFEU, or subject to their application, which relate to the processing of personal data by the Member States when carrying out activities falling within the scope of Chapter 4 or Chapter 5 of Title V of Part Three of the TFEU.

+ (42) In line with the presumption of innocence, natural persons in the Union should always be judged on their actual behaviour. Natural persons should never be judged on AI-predicted behaviour based solely on their profiling, personality traits or characteristics, such as nationality, place of birth, place of residence, number of children, level of debt or type of car, without a reasonable suspicion of that person being involved in a criminal activity based on objective verifiable facts and without human assessment thereof. Therefore, risk assessments carried out with regard to natural persons in order to assess the likelihood of their offending or to predict the occurrence of an actual or potential criminal offence based solely on profiling them or on assessing their personality traits and characteristics should be prohibited. In any case, that prohibition does not refer to or touch upon risk analytics that are not based on the profiling of individuals or on the personality traits and characteristics of individuals, such as AI systems using risk analytics to assess the likelihood of financial fraud by undertakings on the basis of suspicious transactions or risk analytic tools to predict the likelihood of the localisation of narcotics or illicit goods by customs authorities, for example on the basis of known trafficking routes.

+ (43) The placing on the market, the putting into service for that specific purpose, or the use of AI systems that create or expand facial recognition databases through the untargeted scraping of facial images from the internet or CCTV footage, should be prohibited because that practice adds to the feeling of mass surveillance and can lead to gross violations of fundamental rights, including the right to privacy.

+ (44) There are serious concerns about the scientific basis of AI systems aiming to identify or infer emotions, particularly as expression of emotions vary considerably across cultures and situations, and even within a single individual. Among the key shortcomings of such systems are the limited reliability, the lack of specificity and the limited generalisability. Therefore, AI systems identifying or inferring emotions or intentions of natural persons on the basis of their biometric data may lead to discriminatory outcomes and can be intrusive to the rights and freedoms of the concerned persons. Considering the imbalance of power in the context of work or education, combined with the intrusive nature of these systems, such systems could lead to detrimental or unfavourable treatment of certain natural persons or whole groups thereof. Therefore, the placing on the market, the putting into service, or the use of AI systems intended to be used to detect the emotional state of individuals in situations related to the workplace and education should be prohibited. That prohibition should not cover AI systems placed on the market strictly for medical or safety reasons, such as systems intended for therapeutical use.

+ (45) Practices that are prohibited by Union law, including data protection law, non-discrimination law, consumer protection law, and competition law, should not be affected by this Regulation.

+ (46) High-risk AI systems should only be placed on the Union market, put into service or used if they comply with certain mandatory requirements. Those requirements should ensure that high-risk AI systems available in the Union or whose output is otherwise used in the Union do not pose unacceptable risks to important Union public interests as recognised and protected by Union law. On the basis of the New Legislative Framework, as clarified in the Commission notice `The "Blue Guide" on the implementation of EU product rules 2022' (20), the general rule is that more than one legal act of Union harmonisation legislation, such as Regulations (EU) 2017/745 (21) and (EU) 2017/746 (22) of the European Parliament and of the Council or Directive 2006/42/EC of the European Parliament and of the Council (23), may be applicable to one product, since the making available or putting into service can take

+ (20) OJ C 247, 29.6.2022, p. 1. (21) Regulation (EU) 2017/745 of the European Parliament and of the Council of 5 April 2017 on medical devices, amending Directive

+ 2001/83/EC, Regulation (EC) No 178/2002 and Regulation (EC) No 1223/2009 and repealing Council Directives 90/385/EEC and 93/42/EEC (OJ L 117, 5.5.2017, p. 1). (22) Regulation (EU) 2017/746 of the European Parliament and of the Council of 5 April 2017 on in vitro diagnostic medical devices and repealing Directive 98/79/EC and Commission Decision 2010/227/EU (OJ L 117, 5.5.2017, p. 176). (23) Directive 2006/42/EC of the European Parliament and of the Council of 17 May 2006 on machinery, and amending Directive 95/16/EC (OJ L 157, 9.6.2006, p. 24).

+ 12/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ place only when the product complies with all applicable Union harmonisation legislation. To ensure consistency and avoid unnecessary administrative burdens or costs, providers of a product that contains one or more high-risk AI systems, to which the requirements of this Regulation and of the Union harmonisation legislation listed in an annex to this Regulation apply, should have flexibility with regard to operational decisions on how to ensure compliance of a product that contains one or more AI systems with all applicable requirements of the Union harmonisation legislation in an optimal manner. AI systems identified as high-risk should be limited to those that have a significant harmful impact on the health, safety and fundamental rights of persons in the Union and such limitation should minimise any potential restriction to international trade.

+ (47) AI systems could have an adverse impact on the health and safety of persons, in particular when such systems operate as safety components of products. Consistent with the objectives of Union harmonisation legislation to facilitate the free movement of products in the internal market and to ensure that only safe and otherwise compliant products find their way into the market, it is important that the safety risks that may be generated by a product as a whole due to its digital components, including AI systems, are duly prevented and mitigated. For instance, increasingly autonomous robots, whether in the context of manufacturing or personal assistance and care should be able to safely operate and performs their functions in complex environments. Similarly, in the health sector where the stakes for life and health are particularly high, increasingly sophisticated diagnostics systems and systems supporting human decisions should be reliable and accurate.

+ (48) The extent of the adverse impact caused by the AI system on the fundamental rights protected by the Charter is of particular relevance when classifying an AI system as high risk. Those rights include the right to human dignity, respect for private and family life, protection of personal data, freedom of expression and information, freedom of assembly and of association, the right to non-discrimination, the right to education, consumer protection, workers' rights, the rights of persons with disabilities, gender equality, intellectual property rights, the right to an effective remedy and to a fair trial, the right of defence and the presumption of innocence, and the right to good administration. In addition to those rights, it is important to highlight the fact that children have specific rights as enshrined in Article 24 of the Charter and in the United Nations Convention on the Rights of the Child, further developed in the UNCRC General Comment No 25 as regards the digital environment, both of which require consideration of the children's vulnerabilities and provision of such protection and care as necessary for their well-being. The fundamental right to a high level of environmental protection enshrined in the Charter and implemented in Union policies should also be considered when assessing the severity of the harm that an AI system can cause, including in relation to the health and safety of persons.

+ (49) As regards high-risk AI systems that are safety components of products or systems, or which are themselves

+ products or systems falling within the scope of Regulation (EC) No 300/2008 of the European Parliament and of the Council (24), Regulation (EU) No 167/2013 of the European Parliament and of the Council (25), Regulation (EU) No 168/2013 of the European Parliament and of the Council (26), Directive 2014/90/EU of the European Parliament and of the Council (27), Directive (EU) 2016/797 of the European Parliament and of the Council (28), Regulation (EU) 2018/858 of the European Parliament and of the Council (29), Regulation (EU) 2018/1139 of the

+ (24) Regulation (EC) No 300/2008 of the European Parliament and of the Council of 11 March 2008 on common rules in the field of civil aviation security and repealing Regulation (EC) No 2320/2002 (OJ L 97, 9.4.2008, p. 72).

+ (25) Regulation (EU) No 167/2013 of the European Parliament and of the Council of 5 February 2013 on the approval and market surveillance of agricultural and forestry vehicles (OJ L 60, 2.3.2013, p. 1).

+ (26) Regulation (EU) No 168/2013 of the European Parliament and of the Council of 15 January 2013 on the approval and market surveillance of two- or three-wheel vehicles and quadricycles (OJ L 60, 2.3.2013, p. 52).

+ (27) Directive 2014/90/EU of the European Parliament and of the Council of 23 July 2014 on marine equipment and repealing Council Directive 96/98/EC (OJ L 257, 28.8.2014, p. 146).

+ (28) Directive (EU) 2016/797 of the European Parliament and of the Council of 11 May 2016 on the interoperability of the rail system within the European Union (OJ L 138, 26.5.2016, p. 44).

+ (29) Regulation (EU) 2018/858 of the European Parliament and of the Council of 30 May 2018 on the approval and market surveillance of motor vehicles and their trailers, and of systems, components and separate technical units intended for such vehicles, amending Regulations (EC) No 715/2007 and (EC) No 595/2009 and repealing Directive 2007/46/EC (OJ L 151, 14.6.2018, p. 1).

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 13/144

+ European Parliament and of the Council (30), and Regulation (EU) 2019/2144 of the European Parliament and of the Council (31), it is appropriate to amend those acts to ensure that the Commission takes into account, on the basis of the technical and regulatory specificities of each sector, and without interfering with existing governance, conformity assessment and enforcement mechanisms and authorities established therein, the mandatory requirements for high-risk AI systems laid down in this Regulation when adopting any relevant delegated or implementing acts on the basis of those acts.

+ (50) As regards AI systems that are safety components of products, or which are themselves products, falling within the scope of certain Union harmonisation legislation listed in an annex to this Regulation, it is appropriate to classify them as high-risk under this Regulation if the product concerned undergoes the conformity assessment procedure with a third-party conformity assessment body pursuant to that relevant Union harmonisation legislation. In particular, such products are machinery, toys, lifts, equipment and protective systems intended for use in potentially explosive atmospheres, radio equipment, pressure equipment, recreational craft equipment, cableway installations, appliances burning gaseous fuels, medical devices, in vitro diagnostic medical devices, automotive and aviation.

+ (51) The classification of an AI system as high-risk pursuant to this Regulation should not necessarily mean that the product whose safety component is the AI system, or the AI system itself as a product, is considered to be high-risk under the criteria established in the relevant Union harmonisation legislation that applies to the product. This is, in particular, the case for Regulations (EU) 2017/745 and (EU) 2017/746, where a third-party conformity assessment is provided for medium-risk and high-risk products.

+ (52) As regards stand-alone AI systems, namely high-risk AI systems other than those that are safety components of products, or that are themselves products, it is appropriate to classify them as high-risk if, in light of their intended purpose, they pose a high risk of harm to the health and safety or the fundamental rights of persons, taking into account both the severity of the possible harm and its probability of occurrence and they are used in a number of specifically pre-defined areas specified in this Regulation. The identification of those systems is based on the same methodology and criteria envisaged also for any future amendments of the list of high-risk AI systems that the Commission should be empowered to adopt, via delegated acts, to take into account the rapid pace of technological development, as well as the potential changes in the use of AI systems.

+ (53) It is also important to clarify that there may be specific cases in which AI systems referred to in pre-defined areas specified in this Regulation do not lead to a significant risk of harm to the legal interests protected under those areas because they do not materially influence the decision-making or do not harm those interests substantially. For the purposes of this Regulation, an AI system that does not materially influence the outcome of decision-making should be understood to be an AI system that does not have an impact on the substance, and thereby the outcome, of decision-making, whether human or automated. An AI system that does not materially influence the outcome of decision-making could include situations in which one or more of the following conditions are fulfilled. The first such condition should be that the AI system is intended to perform a narrow procedural task, such as an AI system that transforms unstructured data into structured data, an AI system that classifies incoming documents into categories or an AI system that is used to detect duplicates among a large number of applications. Those tasks are of such narrow and limited nature that they pose only limited risks which are not increased through the use of an AI system in a context that is listed as a high-risk use in an annex to this Regulation. The second condition should be

+ (30) Regulation (EU) 2018/1139 of the European Parliament and of the Council of 4 July 2018 on common rules in the field of civil aviation and establishing a European Union Aviation Safety Agency, and amending Regulations (EC) No 2111/2005, (EC) No 1008/2008, (EU) No 996/2010, (EU) No 376/2014 and Directives 2014/30/EU and 2014/53/EU of the European Parliament and of the Council, and repealing Regulations (EC) No 552/2004 and (EC) No 216/2008 of the European Parliament and of the Council and Council Regulation (EEC) No 3922/91 (OJ L 212, 22.8.2018, p. 1).

+ (31) Regulation (EU) 2019/2144 of the European Parliament and of the Council of 27 November 2019 on type-approval requirements for motor vehicles and their trailers, and systems, components and separate technical units intended for such vehicles, as regards their general safety and the protection of vehicle occupants and vulnerable road users, amending Regulation (EU) 2018/858 of the European Parliament and of the Council and repealing Regulations (EC) No 78/2009, (EC) No 79/2009 and (EC) No 661/2009 of the European Parliament and of the Council and Commission Regulations (EC) No 631/2009, (EU) No 406/2010, (EU) No 672/2010, (EU) No 1003/2010, (EU) No 1005/2010, (EU) No 1008/2010, (EU) No 1009/2010, (EU) No 19/2011, (EU) No 109/2011, (EU) No 458/2011, (EU) No 65/2012, (EU) No 130/2012, (EU) No 347/2012, (EU) No 351/2012, (EU) No 1230/2012 and (EU) 2015/166 (OJ L 325, 16.12.2019, p. 1).

+ 14/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ that the task performed by the AI system is intended to improve the result of a previously completed human activity that may be relevant for the purposes of the high-risk uses listed in an annex to this Regulation. Considering those characteristics, the AI system provides only an additional layer to a human activity with consequently lowered risk. That condition would, for example, apply to AI systems that are intended to improve the language used in previously drafted documents, for example in relation to professional tone, academic style of language or by aligning text to a certain brand messaging. The third condition should be that the AI system is intended to detect decision-making patterns or deviations from prior decision-making patterns. The risk would be lowered because the use of the AI system follows a previously completed human assessment which it is not meant to replace or influence, without proper human review. Such AI systems include for instance those that, given a certain grading pattern of a teacher, can be used to check ex post whether the teacher may have deviated from the grading pattern so as to flag potential inconsistencies or anomalies. The fourth condition should be that the AI system is intended to perform a task that is only preparatory to an assessment relevant for the purposes of the AI systems listed in an annex to this Regulation, thus making the possible impact of the output of the system very low in terms of representing a risk for the assessment to follow. That condition covers, inter alia, smart solutions for file handling, which include various functions from indexing, searching, text and speech processing or linking data to other data sources, or AI systems used for translation of initial documents. In any case, AI systems used in high-risk use-cases listed in an annex to this Regulation should be considered to pose significant risks of harm to the health, safety or fundamental rights if the AI system implies profiling within the meaning of Article 4, point (4) of Regulation (EU) 2016/679 or Article 3, point (4) of Directive (EU) 2016/680 or Article 3, point (5) of Regulation (EU) 2018/1725. To ensure traceability and transparency, a provider who considers that an AI system is not high-risk on the basis of the conditions referred to above should draw up documentation of the assessment before that system is placed on the market or put into service and should provide that documentation to national competent authorities upon request. Such a provider should be obliged to register the AI system in the EU database established under this Regulation. With a view to providing further guidance for the practical implementation of the conditions under which the AI systems listed in an annex to this Regulation are, on an exceptional basis, non-high-risk, the Commission should, after consulting the Board, provide guidelines specifying that practical implementation, completed by a comprehensive list of practical examples of use cases of AI systems that are high-risk and use cases that are not.

+ (54) As biometric data constitutes a special category of personal data, it is appropriate to classify as high-risk several critical-use cases of biometric systems, insofar as their use is permitted under relevant Union and national law. Technical inaccuracies of AI systems intended for the remote biometric identification of natural persons can lead to biased results and entail discriminatory effects. The risk of such biased results and discriminatory effects is particularly relevant with regard to age, ethnicity, race, sex or disabilities. Remote biometric identification systems should therefore be classified as high-risk in view of the risks that they pose. Such a classification excludes AI systems intended to be used for biometric verification, including authentication, the sole purpose of which is to confirm that a specific natural person is who that person claims to be and to confirm the identity of a natural person for the sole purpose of having access to a service, unlocking a device or having secure access to premises. In addition, AI systems intended to be used for biometric categorisation according to sensitive attributes or characteristics protected under Article 9(1) of Regulation (EU) 2016/679 on the basis of biometric data, in so far as these are not prohibited under this Regulation, and emotion recognition systems that are not prohibited under this Regulation, should be classified as high-risk. Biometric systems which are intended to be used solely for the purpose of enabling cybersecurity and personal data protection measures should not be considered to be high-risk AI systems.

+ (55) As regards the management and operation of critical infrastructure, it is appropriate to classify as high-risk the AI systems intended to be used as safety components in the management and operation of critical digital infrastructure as listed in point (8) of the Annex to Directive (EU) 2022/2557, road traffic and the supply of water, gas, heating and electricity, since their failure or malfunctioning may put at risk the life and health of persons at large scale and lead to appreciable disruptions in the ordinary conduct of social and economic activities. Safety components of critical infrastructure, including critical digital infrastructure, are systems used to directly protect the physical integrity of critical infrastructure or the health and safety of persons and property but which are not necessary in order for the

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 15/144

+ system to function. The failure or malfunctioning of such components might directly lead to risks to the physical integrity of critical infrastructure and thus to risks to health and safety of persons and property. Components intended to be used solely for cybersecurity purposes should not qualify as safety components. Examples of safety components of such critical infrastructure may include systems for monitoring water pressure or fire alarm controlling systems in cloud computing centres.

+ (56) The deployment of AI systems in education is important to promote high-quality digital education and training and to allow all learners and teachers to acquire and share the necessary digital skills and competences, including media literacy, and critical thinking, to take an active part in the economy, society, and in democratic processes. However, AI systems used in education or vocational training, in particular for determining access or admission, for assigning persons to educational and vocational training institutions or programmes at all levels, for evaluating learning outcomes of persons, for assessing the appropriate level of education for an individual and materially influencing the level of education and training that individuals will receive or will be able to access or for monitoring and detecting prohibited behaviour of students during tests should be classified as high-risk AI systems, since they may determine the educational and professional course of a person's life and therefore may affect that person's ability to secure a livelihood. When improperly designed and used, such systems may be particularly intrusive and may violate the right to education and training as well as the right not to be discriminated against and perpetuate historical patterns of discrimination, for example against women, certain age groups, persons with disabilities, or persons of certain racial or ethnic origins or sexual orientation.

+ (57) AI systems used in employment, workers management and access to self-employment, in particular for the recruitment and selection of persons, for making decisions affecting terms of the work-related relationship, promotion and termination of work-related contractual relationships, for allocating tasks on the basis of individual behaviour, personal traits or characteristics and for monitoring or evaluation of persons in work-related contractual relationships, should also be classified as high-risk, since those systems may have an appreciable impact on future career prospects, livelihoods of those persons and workers' rights. Relevant work-related contractual relationships should, in a meaningful manner, involve employees and persons providing services through platforms as referred to in the Commission Work Programme 2021. Throughout the recruitment process and in the evaluation, promotion, or retention of persons in work-related contractual relationships, such systems may perpetuate historical patterns of discrimination, for example against women, certain age groups, persons with disabilities, or persons of certain racial or ethnic origins or sexual orientation. AI systems used to monitor the performance and behaviour of such persons may also undermine their fundamental rights to data protection and privacy.

+ (58) Another area in which the use of AI systems deserves special consideration is the access to and enjoyment of certain essential private and public services and benefits necessary for people to fully participate in society or to improve one's standard of living. In particular, natural persons applying for or receiving essential public assistance benefits and services from public authorities namely healthcare services, social security benefits, social services providing protection in cases such as maternity, illness, industrial accidents, dependency or old age and loss of employment and social and housing assistance, are typically dependent on those benefits and services and in a vulnerable position in relation to the responsible authorities. If AI systems are used for determining whether such benefits and services should be granted, denied, reduced, revoked or reclaimed by authorities, including whether beneficiaries are legitimately entitled to such benefits or services, those systems may have a significant impact on persons' livelihood and may infringe their fundamental rights, such as the right to social protection, non-discrimination, human dignity or an effective remedy and should therefore be classified as high-risk. Nonetheless, this Regulation should not hamper the development and use of innovative approaches in the public administration, which would stand to benefit from a wider use of compliant and safe AI systems, provided that those systems do not entail a high risk to legal and natural persons. In addition, AI systems used to evaluate the credit score or creditworthiness of natural persons should be classified as high-risk AI systems, since they determine those persons' access to financial resources or essential services such as housing, electricity, and telecommunication services. AI systems used for those purposes may lead to discrimination between persons or groups and may perpetuate historical patterns of discrimination, such as that based on racial or ethnic origins, gender, disabilities, age or sexual orientation, or may create new forms of discriminatory impacts. However, AI systems provided for by Union law for the purpose of detecting fraud in the offering of financial services and for prudential purposes to calculate credit institutions' and insurance undertakings' capital requirements should not be considered to be high-risk under this Regulation. Moreover, AI systems intended

+ 16/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ to be used for risk assessment and pricing in relation to natural persons for health and life insurance can also have a significant impact on persons' livelihood and if not duly designed, developed and used, can infringe their fundamental rights and can lead to serious consequences for people's life and health, including financial exclusion and discrimination. Finally, AI systems used to evaluate and classify emergency calls by natural persons or to dispatch or establish priority in the dispatching of emergency first response services, including by police, firefighters and medical aid, as well as of emergency healthcare patient triage systems, should also be classified as high-risk since they make decisions in very critical situations for the life and health of persons and their property.

+ (59) Given their role and responsibility, actions by law enforcement authorities involving certain uses of AI systems are characterised by a significant degree of power imbalance and may lead to surveillance, arrest or deprivation of a natural person's liberty as well as other adverse impacts on fundamental rights guaranteed in the Charter. In particular, if the AI system is not trained with high-quality data, does not meet adequate requirements in terms of its performance, its accuracy or robustness, or is not properly designed and tested before being put on the market or otherwise put into service, it may single out people in a discriminatory or otherwise incorrect or unjust manner. Furthermore, the exercise of important procedural fundamental rights, such as the right to an effective remedy and to a fair trial as well as the right of defence and the presumption of innocence, could be hampered, in particular, where such AI systems are not sufficiently transparent, explainable and documented. It is therefore appropriate to classify as high-risk, insofar as their use is permitted under relevant Union and national law, a number of AI systems intended to be used in the law enforcement context where accuracy, reliability and transparency is particularly important to avoid adverse impacts, retain public trust and ensure accountability and effective redress. In view of the nature of the activities and the risks relating thereto, those high-risk AI systems should include in particular AI systems intended to be used by or on behalf of law enforcement authorities or by Union institutions, bodies, offices, or agencies in support of law enforcement authorities for assessing the risk of a natural person to become a victim of criminal offences, as polygraphs and similar tools, for the evaluation of the reliability of evidence in in the course of investigation or prosecution of criminal offences, and, insofar as not prohibited under this Regulation, for assessing the risk of a natural person offending or reoffending not solely on the basis of the profiling of natural persons or the assessment of personality traits and characteristics or the past criminal behaviour of natural persons or groups, for profiling in the course of detection, investigation or prosecution of criminal offences. AI systems specifically intended to be used for administrative proceedings by tax and customs authorities as well as by financial intelligence units carrying out administrative tasks analysing information pursuant to Union anti-money laundering law should not be classified as high-risk AI systems used by law enforcement authorities for the purpose of prevention, detection, investigation and prosecution of criminal offences. The use of AI tools by law enforcement and other relevant authorities should not become a factor of inequality, or exclusion. The impact of the use of AI tools on the defence rights of suspects should not be ignored, in particular the difficulty in obtaining meaningful information on the functioning of those systems and the resulting difficulty in challenging their results in court, in particular by natural persons under investigation.

+ (60) AI systems used in migration, asylum and border control management affect persons who are often in particularly vulnerable position and who are dependent on the outcome of the actions of the competent public authorities. The accuracy, non-discriminatory nature and transparency of the AI systems used in those contexts are therefore particularly important to guarantee respect for the fundamental rights of the affected persons, in particular their rights to free movement, non-discrimination, protection of private life and personal data, international protection and good administration. It is therefore appropriate to classify as high-risk, insofar as their use is permitted under relevant Union and national law, AI systems intended to be used by or on behalf of competent public authorities or by Union institutions, bodies, offices or agencies charged with tasks in the fields of migration, asylum and border control management as polygraphs and similar tools, for assessing certain risks posed by natural persons entering the territory of a Member State or applying for visa or asylum, for assisting competent public authorities for the examination, including related assessment of the reliability of evidence, of applications for asylum, visa and residence permits and associated complaints with regard to the objective to establish the eligibility of the natural persons applying for a status, for the purpose of detecting, recognising or identifying natural persons in the context of migration, asylum and border control management, with the exception of verification of travel documents. AI systems in the area of migration, asylum and border control management covered by this Regulation should comply with the relevant procedural requirements set by the Regulation (EC) No 810/2009 of the European Parliament and

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 17/144

+ of the Council (32), the Directive 2013/32/EU of the European Parliament and of the Council (33), and other relevant Union law. The use of AI systems in migration, asylum and border control management should, in no circumstances, be used by Member States or Union institutions, bodies, offices or agencies as a means to circumvent their international obligations under the UN Convention relating to the Status of Refugees done at Geneva on 28 July 1951 as amended by the Protocol of 31 January 1967. Nor should they be used to in any way infringe on the principle of non-refoulement, or to deny safe and effective legal avenues into the territory of the Union, including the right to international protection.

+ (61) Certain AI systems intended for the administration of justice and democratic processes should be classified as high-risk, considering their potentially significant impact on democracy, the rule of law, individual freedoms as well as the right to an effective remedy and to a fair trial. In particular, to address the risks of potential biases, errors and opacity, it is appropriate to qualify as high-risk AI systems intended to be used by a judicial authority or on its behalf to assist judicial authorities in researching and interpreting facts and the law and in applying the law to a concrete set of facts. AI systems intended to be used by alternative dispute resolution bodies for those purposes should also be considered to be high-risk when the outcomes of the alternative dispute resolution proceedings produce legal effects for the parties. The use of AI tools can support the decision-making power of judges or judicial independence, but should not replace it: the final decision-making must remain a human-driven activity. The classification of AI systems as high-risk should not, however, extend to AI systems intended for purely ancillary administrative activities that do not affect the actual administration of justice in individual cases, such as anonymisation or pseudonymisation of judicial decisions, documents or data, communication between personnel, administrative tasks.

+ (62) Without prejudice to the rules provided for in Regulation (EU) 2024/900 of the European Parliament and of the Council (34), and in order to address the risks of undue external interference with the right to vote enshrined in Article 39 of the Charter, and of adverse effects on democracy and the rule of law, AI systems intended to be used to influence the outcome of an election or referendum or the voting behaviour of natural persons in the exercise of their vote in elections or referenda should be classified as high-risk AI systems with the exception of AI systems whose output natural persons are not directly exposed to, such as tools used to organise, optimise and structure political campaigns from an administrative and logistical point of view.

+ (63) The fact that an AI system is classified as a high-risk AI system under this Regulation should not be interpreted as indicating that the use of the system is lawful under other acts of Union law or under national law compatible with Union law, such as on the protection of personal data, on the use of polygraphs and similar tools or other systems to detect the emotional state of natural persons. Any such use should continue to occur solely in accordance with the applicable requirements resulting from the Charter and from the applicable acts of secondary Union law and national law. This Regulation should not be understood as providing for the legal ground for processing of personal data, including special categories of personal data, where relevant, unless it is specifically otherwise provided for in this Regulation.

+ (64) To mitigate the risks from high-risk AI systems placed on the market or put into service and to ensure a high level of trustworthiness, certain mandatory requirements should apply to high-risk AI systems, taking into account the intended purpose and the context of use of the AI system and according to the risk-management system to be established by the provider. The measures adopted by the providers to comply with the mandatory requirements of this Regulation should take into account the generally acknowledged state of the art on AI, be proportionate and effective to meet the objectives of this Regulation. Based on the New Legislative Framework, as clarified in Commission notice `The "Blue Guide" on the implementation of EU product rules 2022', the general rule is that more than one legal act of Union harmonisation legislation may be applicable to one product, since the making available or putting into service can take place only when the product complies with all applicable Union harmonisation legislation. The hazards of AI systems covered by the requirements of this Regulation concern different aspects than the existing Union harmonisation legislation and therefore the requirements of this Regulation would complement the existing body of the Union harmonisation legislation. For example, machinery or medical devices products incorporating an AI system might present risks not addressed by the essential health and safety

+ (32) Regulation (EC) No 810/2009 of the European Parliament and of the Council of 13 July 2009 establishing a Community Code on Visas (Visa Code) (OJ L 243, 15.9.2009, p. 1).

+ (33) Directive 2013/32/EU of the European Parliament and of the Council of 26 June 2013 on common procedures for granting and withdrawing international protection (OJ L 180, 29.6.2013, p. 60).

+ (34) Regulation (EU) 2024/900 of the European parliament and of the Council of 13 March 2024 on the transparency and targeting of political advertising (OJ L, 2024/900, 20.3.2024, ELI: http://data.europa.eu/eli/reg/2024/900/oj).

+ 18/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ requirements set out in the relevant Union harmonised legislation, as that sectoral law does not deal with risks specific to AI systems. This calls for a simultaneous and complementary application of the various legislative acts. To ensure consistency and to avoid an unnecessary administrative burden and unnecessary costs, providers of a product that contains one or more high-risk AI system, to which the requirements of this Regulation and of the Union harmonisation legislation based on the New Legislative Framework and listed in an annex to this Regulation apply, should have flexibility with regard to operational decisions on how to ensure compliance of a product that contains one or more AI systems with all the applicable requirements of that Union harmonised legislation in an optimal manner. That flexibility could mean, for example a decision by the provider to integrate a part of the necessary testing and reporting processes, information and documentation required under this Regulation into already existing documentation and procedures required under existing Union harmonisation legislation based on the New Legislative Framework and listed in an annex to this Regulation. This should not, in any way, undermine the obligation of the provider to comply with all the applicable requirements.

+ (65) The risk-management system should consist of a continuous, iterative process that is planned and run throughout the entire lifecycle of a high-risk AI system. That process should be aimed at identifying and mitigating the relevant risks of AI systems on health, safety and fundamental rights. The risk-management system should be regularly reviewed and updated to ensure its continuing effectiveness, as well as justification and documentation of any significant decisions and actions taken subject to this Regulation. This process should ensure that the provider identifies risks or adverse impacts and implements mitigation measures for the known and reasonably foreseeable risks of AI systems to the health, safety and fundamental rights in light of their intended purpose and reasonably foreseeable misuse, including the possible risks arising from the interaction between the AI system and the environment within which it operates. The risk-management system should adopt the most appropriate risk-management measures in light of the state of the art in AI. When identifying the most appropriate risk-management measures, the provider should document and explain the choices made and, when relevant, involve experts and external stakeholders. In identifying the reasonably foreseeable misuse of high-risk AI systems, the provider should cover uses of AI systems which, while not directly covered by the intended purpose and provided for in the instruction for use may nevertheless be reasonably expected to result from readily predictable human behaviour in the context of the specific characteristics and use of a particular AI system. Any known or foreseeable circumstances related to the use of the high-risk AI system in accordance with its intended purpose or under conditions of reasonably foreseeable misuse, which may lead to risks to the health and safety or fundamental rights should be included in the instructions for use that are provided by the provider. This is to ensure that the deployer is aware and takes them into account when using the high-risk AI system. Identifying and implementing risk mitigation measures for foreseeable misuse under this Regulation should not require specific additional training for the high-risk AI system by the provider to address foreseeable misuse. The providers however are encouraged to consider such additional training measures to mitigate reasonable foreseeable misuses as necessary and appropriate.

+ (66) Requirements should apply to high-risk AI systems as regards risk management, the quality and relevance of data sets used, technical documentation and record-keeping, transparency and the provision of information to deployers, human oversight, and robustness, accuracy and cybersecurity. Those requirements are necessary to effectively mitigate the risks for health, safety and fundamental rights. As no other less trade restrictive measures are reasonably available those requirements are not unjustified restrictions to trade.

+ (67) High-quality data and access to high-quality data plays a vital role in providing structure and in ensuring the performance of many AI systems, especially when techniques involving the training of models are used, with a view to ensure that the high-risk AI system performs as intended and safely and it does not become a source of discrimination prohibited by Union law. High-quality data sets for training, validation and testing require the implementation of appropriate data governance and management practices. Data sets for training, validation and testing, including the labels, should be relevant, sufficiently representative, and to the best extent possible free of errors and complete in view of the intended purpose of the system. In order to facilitate compliance with Union data protection law, such as Regulation (EU) 2016/679, data governance and management practices should include, in the case of personal data, transparency about the original purpose of the data collection. The data sets should also have the appropriate statistical properties, including as regards the persons or groups of persons in relation to whom the high-risk AI system is intended to be used, with specific attention to the mitigation of possible biases in the data sets, that are likely to affect the health and safety of persons, have a negative impact on fundamental rights or lead to discrimination prohibited under Union law, especially where data outputs influence inputs for future operations

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 19/144

+ (feedback loops). Biases can for example be inherent in underlying data sets, especially when historical data is being used, or generated when the systems are implemented in real world settings. Results provided by AI systems could be influenced by such inherent biases that are inclined to gradually increase and thereby perpetuate and amplify existing discrimination, in particular for persons belonging to certain vulnerable groups, including racial or ethnic groups. The requirement for the data sets to be to the best extent possible complete and free of errors should not affect the use of privacy-preserving techniques in the context of the development and testing of AI systems. In particular, data sets should take into account, to the extent required by their intended purpose, the features, characteristics or elements that are particular to the specific geographical, contextual, behavioural or functional setting which the AI system is intended to be used. The requirements related to data governance can be complied with by having recourse to third parties that offer certified compliance services including verification of data governance, data set integrity, and data training, validation and testing practices, as far as compliance with the data requirements of this Regulation are ensured.

+ (68) For the development and assessment of high-risk AI systems, certain actors, such as providers, notified bodies and other relevant entities, such as European Digital Innovation Hubs, testing experimentation facilities and researchers, should be able to access and use high-quality data sets within the fields of activities of those actors which are related to this Regulation. European common data spaces established by the Commission and the facilitation of data sharing between businesses and with government in the public interest will be instrumental to provide trustful, accountable and non-discriminatory access to high-quality data for the training, validation and testing of AI systems. For example, in health, the European health data space will facilitate non-discriminatory access to health data and the training of AI algorithms on those data sets, in a privacy-preserving, secure, timely, transparent and trustworthy manner, and with an appropriate institutional governance. Relevant competent authorities, including sectoral ones, providing or supporting the access to data may also support the provision of high-quality data for the training, validation and testing of AI systems.

+ (69) The right to privacy and to protection of personal data must be guaranteed throughout the entire lifecycle of the AI system. In this regard, the principles of data minimisation and data protection by design and by default, as set out in Union data protection law, are applicable when personal data are processed. Measures taken by providers to ensure compliance with those principles may include not only anonymisation and encryption, but also the use of technology that permits algorithms to be brought to the data and allows training of AI systems without the transmission between parties or copying of the raw or structured data themselves, without prejudice to the requirements on data governance provided for in this Regulation.

+ (70) In order to protect the right of others from the discrimination that might result from the bias in AI systems, the providers should, exceptionally, to the extent that it is strictly necessary for the purpose of ensuring bias detection and correction in relation to the high-risk AI systems, subject to appropriate safeguards for the fundamental rights and freedoms of natural persons and following the application of all applicable conditions laid down under this Regulation in addition to the conditions laid down in Regulations (EU) 2016/679 and (EU) 2018/1725 and Directive (EU) 2016/680, be able to process also special categories of personal data, as a matter of substantial public interest within the meaning of Article 9(2), point (g) of Regulation (EU) 2016/679 and Article 10(2), point (g) of Regulation (EU) 2018/1725.

+ (71) Having comprehensible information on how high-risk AI systems have been developed and how they perform throughout their lifetime is essential to enable traceability of those systems, verify compliance with the requirements under this Regulation, as well as monitoring of their operations and post market monitoring. This requires keeping records and the availability of technical documentation, containing information which is necessary to assess the compliance of the AI system with the relevant requirements and facilitate post market monitoring. Such information should include the general characteristics, capabilities and limitations of the system, algorithms, data, training, testing and validation processes used as well as documentation on the relevant risk-management system and drawn in a clear and comprehensive form. The technical documentation should be kept up to date, appropriately throughout the lifetime of the AI system. Furthermore, high-risk AI systems should technically allow for the automatic recording of events, by means of logs, over the duration of the lifetime of the system.

+ 20/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ (72) To address concerns related to opacity and complexity of certain AI systems and help deployers to fulfil their obligations under this Regulation, transparency should be required for high-risk AI systems before they are placed on the market or put it into service. High-risk AI systems should be designed in a manner to enable deployers to understand how the AI system works, evaluate its functionality, and comprehend its strengths and limitations. High-risk AI systems should be accompanied by appropriate information in the form of instructions of use. Such information should include the characteristics, capabilities and limitations of performance of the AI system. Those would cover information on possible known and foreseeable circumstances related to the use of the high-risk AI system, including deployer action that may influence system behaviour and performance, under which the AI system can lead to risks to health, safety, and fundamental rights, on the changes that have been pre-determined and assessed for conformity by the provider and on the relevant human oversight measures, including the measures to facilitate the interpretation of the outputs of the AI system by the deployers. Transparency, including the accompanying instructions for use, should assist deployers in the use of the system and support informed decision making by them. Deployers should, inter alia, be in a better position to make the correct choice of the system that they intend to use in light of the obligations applicable to them, be educated about the intended and precluded uses, and use the AI system correctly and as appropriate. In order to enhance legibility and accessibility of the information included in the instructions of use, where appropriate, illustrative examples, for instance on the limitations and on the intended and precluded uses of the AI system, should be included. Providers should ensure that all documentation, including the instructions for use, contains meaningful, comprehensive, accessible and understandable information, taking into account the needs and foreseeable knowledge of the target deployers. Instructions for use should be made available in a language which can be easily understood by target deployers, as determined by the Member State concerned.

+ (73) High-risk AI systems should be designed and developed in such a way that natural persons can oversee their functioning, ensure that they are used as intended and that their impacts are addressed over the system's lifecycle. To that end, appropriate human oversight measures should be identified by the provider of the system before its placing on the market or putting into service. In particular, where appropriate, such measures should guarantee that the system is subject to in-built operational constraints that cannot be overridden by the system itself and is responsive to the human operator, and that the natural persons to whom human oversight has been assigned have the necessary competence, training and authority to carry out that role. It is also essential, as appropriate, to ensure that high-risk AI systems include mechanisms to guide and inform a natural person to whom human oversight has been assigned to make informed decisions if, when and how to intervene in order to avoid negative consequences or risks, or stop the system if it does not perform as intended. Considering the significant consequences for persons in the case of an incorrect match by certain biometric identification systems, it is appropriate to provide for an enhanced human oversight requirement for those systems so that no action or decision may be taken by the deployer on the basis of the identification resulting from the system unless this has been separately verified and confirmed by at least two natural persons. Those persons could be from one or more entities and include the person operating or using the system. This requirement should not pose unnecessary burden or delays and it could be sufficient that the separate verifications by the different persons are automatically recorded in the logs generated by the system. Given the specificities of the areas of law enforcement, migration, border control and asylum, this requirement should not apply where Union or national law considers the application of that requirement to be disproportionate.

+ (74) High-risk AI systems should perform consistently throughout their lifecycle and meet an appropriate level of accuracy, robustness and cybersecurity, in light of their intended purpose and in accordance with the generally acknowledged state of the art. The Commission and relevant organisations and stakeholders are encouraged to take due consideration of the mitigation of risks and the negative impacts of the AI system. The expected level of performance metrics should be declared in the accompanying instructions of use. Providers are urged to communicate that information to deployers in a clear and easily understandable way, free of misunderstandings or misleading statements. Union law on legal metrology, including Directives 2014/31/EU (35) and 2014/32/EU (36) of the European Parliament and of the Council, aims to ensure the accuracy of measurements and to help the transparency and fairness of commercial transactions. In that context, in cooperation with relevant stakeholders and organisation, such as metrology and benchmarking authorities, the Commission should encourage, as appropriate, the development of benchmarks and measurement methodologies for AI systems. In doing so, the Commission should take note and collaborate with international partners working on metrology and relevant measurement indicators relating to AI.

+ (35) Directive 2014/31/EU of the European Parliament and of the Council of 26 February 2014 on the harmonisation of the laws of the Member States relating to the making available on the market of non-automatic weighing instruments (OJ L 96, 29.3.2014, p. 107).

+ (36) Directive 2014/32/EU of the European Parliament and of the Council of 26 February 2014 on the harmonisation of the laws of the Member States relating to the making available on the market of measuring instruments (OJ L 96, 29.3.2014, p. 149).

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 21/144

+ (75) Technical robustness is a key requirement for high-risk AI systems. They should be resilient in relation to harmful or otherwise undesirable behaviour that may result from limitations within the systems or the environment in which the systems operate (e.g. errors, faults, inconsistencies, unexpected situations). Therefore, technical and organisational measures should be taken to ensure robustness of high-risk AI systems, for example by designing and developing appropriate technical solutions to prevent or minimise harmful or otherwise undesirable behaviour. Those technical solution may include for instance mechanisms enabling the system to safely interrupt its operation (fail-safe plans) in the presence of certain anomalies or when operation takes place outside certain predetermined boundaries. Failure to protect against these risks could lead to safety impacts or negatively affect the fundamental rights, for example due to erroneous decisions or wrong or biased outputs generated by the AI system.

+ (76) Cybersecurity plays a crucial role in ensuring that AI systems are resilient against attempts to alter their use, behaviour, performance or compromise their security properties by malicious third parties exploiting the system's vulnerabilities. Cyberattacks against AI systems can leverage AI specific assets, such as training data sets (e.g. data poisoning) or trained models (e.g. adversarial attacks or membership inference), or exploit vulnerabilities in the AI system's digital assets or the underlying ICT infrastructure. To ensure a level of cybersecurity appropriate to the risks, suitable measures, such as security controls, should therefore be taken by the providers of high-risk AI systems, also taking into account as appropriate the underlying ICT infrastructure.

+ (77) Without prejudice to the requirements related to robustness and accuracy set out in this Regulation, high-risk AI systems which fall within the scope of a regulation of the European Parliament and of the Council on horizontal cybersecurity requirements for products with digital elements, in accordance with that regulation may demonstrate compliance with the cybersecurity requirements of this Regulation by fulfilling the essential cybersecurity requirements set out in that regulation. When high-risk AI systems fulfil the essential requirements of a regulation of the European Parliament and of the Council on horizontal cybersecurity requirements for products with digital elements, they should be deemed compliant with the cybersecurity requirements set out in this Regulation in so far as the achievement of those requirements is demonstrated in the EU declaration of conformity or parts thereof issued under that regulation. To that end, the assessment of the cybersecurity risks, associated to a product with digital elements classified as high-risk AI system according to this Regulation, carried out under a regulation of the European Parliament and of the Council on horizontal cybersecurity requirements for products with digital elements, should consider risks to the cyber resilience of an AI system as regards attempts by unauthorised third parties to alter its use, behaviour or performance, including AI specific vulnerabilities such as data poisoning or adversarial attacks, as well as, as relevant, risks to fundamental rights as required by this Regulation.

+ (78) The conformity assessment procedure provided by this Regulation should apply in relation to the essential cybersecurity requirements of a product with digital elements covered by a regulation of the European Parliament and of the Council on horizontal cybersecurity requirements for products with digital elements and classified as a high-risk AI system under this Regulation. However, this rule should not result in reducing the necessary level of assurance for critical products with digital elements covered by a regulation of the European Parliament and of the Council on horizontal cybersecurity requirements for products with digital elements. Therefore, by way of derogation from this rule, high-risk AI systems that fall within the scope of this Regulation and are also qualified as important and critical products with digital elements pursuant to a regulation of the European Parliament and of the Council on horizontal cybersecurity requirements for products with digital elements and to which the conformity assessment procedure based on internal control set out in an annex to this Regulation applies, are subject to the conformity assessment provisions of a regulation of the European Parliament and of the Council on horizontal cybersecurity requirements for products with digital elements insofar as the essential cybersecurity requirements of that regulation are concerned. In this case, for all the other aspects covered by this Regulation the respective provisions on conformity assessment based on internal control set out in an annex to this Regulation should apply. Building on the knowledge and expertise of ENISA on the cybersecurity policy and tasks assigned to ENISA under the Regulation (EU) 2019/881 of the European Parliament and of the Council (37), the Commission should cooperate with ENISA on issues related to cybersecurity of AI systems.

+ (37) Regulation (EU) 2019/881 of the European Parliament and of the Council of 17 April 2019 on ENISA (the European Union Agency for Cybersecurity) and on information and communications technology cybersecurity certification and repealing Regulation (EU) No 526/2013 (Cybersecurity Act) (OJ L 151, 7.6.2019, p. 15).

+ 22/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ (79) It is appropriate that a specific natural or legal person, defined as the provider, takes responsibility for the placing on the market or the putting into service of a high-risk AI system, regardless of whether that natural or legal person is the person who designed or developed the system.

+ (80) As signatories to the United Nations Convention on the Rights of Persons with Disabilities, the Union and the Member States are legally obliged to protect persons with disabilities from discrimination and promote their equality, to ensure that persons with disabilities have access, on an equal basis with others, to information and communications technologies and systems, and to ensure respect for privacy for persons with disabilities. Given the growing importance and use of AI systems, the application of universal design principles to all new technologies and services should ensure full and equal access for everyone potentially affected by or using AI technologies, including persons with disabilities, in a way that takes full account of their inherent dignity and diversity. It is therefore essential that providers ensure full compliance with accessibility requirements, including Directive (EU) 2016/2102 of the European Parliament and of the Council (38) and Directive (EU) 2019/882. Providers should ensure compliance with these requirements by design. Therefore, the necessary measures should be integrated as much as possible into the design of the high-risk AI system.

+ (81) The provider should establish a sound quality management system, ensure the accomplishment of the required conformity assessment procedure, draw up the relevant documentation and establish a robust post-market monitoring system. Providers of high-risk AI systems that are subject to obligations regarding quality management systems under relevant sectoral Union law should have the possibility to include the elements of the quality management system provided for in this Regulation as part of the existing quality management system provided for in that other sectoral Union law. The complementarity between this Regulation and existing sectoral Union law should also be taken into account in future standardisation activities or guidance adopted by the Commission. Public authorities which put into service high-risk AI systems for their own use may adopt and implement the rules for the quality management system as part of the quality management system adopted at a national or regional level, as appropriate, taking into account the specificities of the sector and the competences and organisation of the public authority concerned.

+ (82) To enable enforcement of this Regulation and create a level playing field for operators, and, taking into account the different forms of making available of digital products, it is important to ensure that, under all circumstances, a person established in the Union can provide authorities with all the necessary information on the compliance of an AI system. Therefore, prior to making their AI systems available in the Union, providers established in third countries should, by written mandate, appoint an authorised representative established in the Union. This authorised representative plays a pivotal role in ensuring the compliance of the high-risk AI systems placed on the market or put into service in the Union by those providers who are not established in the Union and in serving as their contact person established in the Union.

+ (83) In light of the nature and complexity of the value chain for AI systems and in line with the New Legislative Framework, it is essential to ensure legal certainty and facilitate the compliance with this Regulation. Therefore, it is necessary to clarify the role and the specific obligations of relevant operators along that value chain, such as importers and distributors who may contribute to the development of AI systems. In certain situations those operators could act in more than one role at the same time and should therefore fulfil cumulatively all relevant obligations associated with those roles. For example, an operator could act as a distributor and an importer at the same time.

+ (84) To ensure legal certainty, it is necessary to clarify that, under certain specific conditions, any distributor, importer, deployer or other third-party should be considered to be a provider of a high-risk AI system and therefore assume all the relevant obligations. This would be the case if that party puts its name or trademark on a high-risk AI system already placed on the market or put into service, without prejudice to contractual arrangements stipulating that the obligations are allocated otherwise. This would also be the case if that party makes a substantial modification to a high-risk AI system that has already been placed on the market or has already been put into service in a way that it remains a high-risk AI system in accordance with this Regulation, or if it modifies the intended purpose of an AI system, including a general-purpose AI system, which has not been classified as high-risk and has already been placed on the market or put into service, in a way that the AI system becomes a high-risk AI system in accordance with this Regulation. Those provisions should apply without prejudice to more specific provisions established in certain Union harmonisation legislation based on the New Legislative Framework, together with which this

+ (38) Directive (EU) 2016/2102 of the European Parliament and of the Council of 26 October 2016 on the accessibility of the websites and mobile applications of public sector bodies (OJ L 327, 2.12.2016, p. 1).

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 23/144

+ Regulation should apply. For example, Article 16(2) of Regulation (EU) 2017/745, establishing that certain changes should not be considered to be modifications of a device that could affect its compliance with the applicable requirements, should continue to apply to high-risk AI systems that are medical devices within the meaning of that Regulation.

+ (85) General-purpose AI systems may be used as high-risk AI systems by themselves or be components of other high-risk AI systems. Therefore, due to their particular nature and in order to ensure a fair sharing of responsibilities along the AI value chain, the providers of such systems should, irrespective of whether they may be used as high-risk AI systems as such by other providers or as components of high-risk AI systems and unless provided otherwise under this Regulation, closely cooperate with the providers of the relevant high-risk AI systems to enable their compliance with the relevant obligations under this Regulation and with the competent authorities established under this Regulation.

+ (86) Where, under the conditions laid down in this Regulation, the provider that initially placed the AI system on the market or put it into service should no longer be considered to be the provider for the purposes of this Regulation, and when that provider has not expressly excluded the change of the AI system into a high-risk AI system, the former provider should nonetheless closely cooperate and make available the necessary information and provide the reasonably expected technical access and other assistance that are required for the fulfilment of the obligations set out in this Regulation, in particular regarding the compliance with the conformity assessment of high-risk AI systems.

+ (87) In addition, where a high-risk AI system that is a safety component of a product which falls within the scope of Union harmonisation legislation based on the New Legislative Framework is not placed on the market or put into service independently from the product, the product manufacturer defined in that legislation should comply with the obligations of the provider established in this Regulation and should, in particular, ensure that the AI system embedded in the final product complies with the requirements of this Regulation.

+ (88) Along the AI value chain multiple parties often supply AI systems, tools and services but also components or processes that are incorporated by the provider into the AI system with various objectives, including the model training, model retraining, model testing and evaluation, integration into software, or other aspects of model development. Those parties have an important role to play in the value chain towards the provider of the high-risk AI system into which their AI systems, tools, services, components or processes are integrated, and should provide by written agreement this provider with the necessary information, capabilities, technical access and other assistance based on the generally acknowledged state of the art, in order to enable the provider to fully comply with the obligations set out in this Regulation, without compromising their own intellectual property rights or trade secrets.

+ (89) Third parties making accessible to the public tools, services, processes, or AI components other than general-purpose AI models, should not be mandated to comply with requirements targeting the responsibilities along the AI value chain, in particular towards the provider that has used or integrated them, when those tools, services, processes, or AI components are made accessible under a free and open-source licence. Developers of free and open-source tools, services, processes, or AI components other than general-purpose AI models should be encouraged to implement widely adopted documentation practices, such as model cards and data sheets, as a way to accelerate information sharing along the AI value chain, allowing the promotion of trustworthy AI systems in the Union.

+ (90) The Commission could develop and recommend voluntary model contractual terms between providers of high-risk AI systems and third parties that supply tools, services, components or processes that are used or integrated in high-risk AI systems, to facilitate the cooperation along the value chain. When developing voluntary model contractual terms, the Commission should also take into account possible contractual requirements applicable in specific sectors or business cases.

+ (91) Given the nature of AI systems and the risks to safety and fundamental rights possibly associated with their use, including as regards the need to ensure proper monitoring of the performance of an AI system in a real-life setting, it is appropriate to set specific responsibilities for deployers. Deployers should in particular take appropriate technical and organisational measures to ensure they use high-risk AI systems in accordance with the instructions of use and certain other obligations should be provided for with regard to monitoring of the functioning of the AI systems and with regard to record-keeping, as appropriate. Furthermore, deployers should ensure that the persons assigned to implement the instructions for use and human oversight as set out in this Regulation have the necessary

+ 24/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ competence, in particular an adequate level of AI literacy, training and authority to properly fulfil those tasks. Those obligations should be without prejudice to other deployer obligations in relation to high-risk AI systems under Union or national law.

+ (92) This Regulation is without prejudice to obligations for employers to inform or to inform and consult workers or their representatives under Union or national law and practice, including Directive 2002/14/EC of the European Parliament and of the Council (39), on decisions to put into service or use AI systems. It remains necessary to ensure information of workers and their representatives on the planned deployment of high-risk AI systems at the workplace where the conditions for those information or information and consultation obligations in other legal instruments are not fulfilled. Moreover, such information right is ancillary and necessary to the objective of protecting fundamental rights that underlies this Regulation. Therefore, an information requirement to that effect should be laid down in this Regulation, without affecting any existing rights of workers.

+ (93) Whilst risks related to AI systems can result from the way such systems are designed, risks can as well stem from how such AI systems are used. Deployers of high-risk AI system therefore play a critical role in ensuring that fundamental rights are protected, complementing the obligations of the provider when developing the AI system. Deployers are best placed to understand how the high-risk AI system will be used concretely and can therefore identify potential significant risks that were not foreseen in the development phase, due to a more precise knowledge of the context of use, the persons or groups of persons likely to be affected, including vulnerable groups. Deployers of high-risk AI systems listed in an annex to this Regulation also play a critical role in informing natural persons and should, when they make decisions or assist in making decisions related to natural persons, where applicable, inform the natural persons that they are subject to the use of the high-risk AI system. This information should include the intended purpose and the type of decisions it makes. The deployer should also inform the natural persons about their right to an explanation provided under this Regulation. With regard to high-risk AI systems used for law enforcement purposes, that obligation should be implemented in accordance with Article 13 of Directive (EU) 2016/680.

+ (94) Any processing of biometric data involved in the use of AI systems for biometric identification for the purpose of law enforcement needs to comply with Article 10 of Directive (EU) 2016/680, that allows such processing only where strictly necessary, subject to appropriate safeguards for the rights and freedoms of the data subject, and where authorised by Union or Member State law. Such use, when authorised, also needs to respect the principles laid down in Article 4 (1) of Directive (EU) 2016/680 including lawfulness, fairness and transparency, purpose limitation, accuracy and storage limitation.

+ (95) Without prejudice to applicable Union law, in particular Regulation (EU) 2016/679 and Directive (EU) 2016/680, considering the intrusive nature of post-remote biometric identification systems, the use of post-remote biometric identification systems should be subject to safeguards. Post-remote biometric identification systems should always be used in a way that is proportionate, legitimate and strictly necessary, and thus targeted, in terms of the individuals to be identified, the location, temporal scope and based on a closed data set of legally acquired video footage. In any case, post-remote biometric identification systems should not be used in the framework of law enforcement to lead to indiscriminate surveillance. The conditions for post-remote biometric identification should in any case not provide a basis to circumvent the conditions of the prohibition and strict exceptions for real time remote biometric identification.

+ (96) In order to efficiently ensure that fundamental rights are protected, deployers of high-risk AI systems that are bodies governed by public law, or private entities providing public services and deployers of certain high-risk AI systems listed in an annex to this Regulation, such as banking or insurance entities, should carry out a fundamental rights impact assessment prior to putting it into use. Services important for individuals that are of public nature may also be provided by private entities. Private entities providing such public services are linked to tasks in the public interest such as in the areas of education, healthcare, social services, housing, administration of justice. The aim of the fundamental rights impact assessment is for the deployer to identify the specific risks to the rights of individuals or groups of individuals likely to be affected, identify measures to be taken in the case of a materialisation of those risks. The impact assessment should be performed prior to deploying the high-risk AI system, and should be updated

+ (39) Directive 2002/14/EC of the European Parliament and of the Council of 11 March 2002 establishing a general framework for informing and consulting employees in the European Community (OJ L 80, 23.3.2002, p. 29).

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 25/144

+ when the deployer considers that any of the relevant factors have changed. The impact assessment should identify the deployer's relevant processes in which the high-risk AI system will be used in line with its intended purpose, and should include a description of the period of time and frequency in which the system is intended to be used as well as of specific categories of natural persons and groups who are likely to be affected in the specific context of use. The assessment should also include the identification of specific risks of harm likely to have an impact on the fundamental rights of those persons or groups. While performing this assessment, the deployer should take into account information relevant to a proper assessment of the impact, including but not limited to the information given by the provider of the high-risk AI system in the instructions for use. In light of the risks identified, deployers should determine measures to be taken in the case of a materialisation of those risks, including for example governance arrangements in that specific context of use, such as arrangements for human oversight according to the instructions of use or, complaint handling and redress procedures, as they could be instrumental in mitigating risks to fundamental rights in concrete use-cases. After performing that impact assessment, the deployer should notify the relevant market surveillance authority. Where appropriate, to collect relevant information necessary to perform the impact assessment, deployers of high-risk AI system, in particular when AI systems are used in the public sector, could involve relevant stakeholders, including the representatives of groups of persons likely to be affected by the AI system, independent experts, and civil society organisations in conducting such impact assessments and designing measures to be taken in the case of materialisation of the risks. The European Artificial Intelligence Office (AI Office) should develop a template for a questionnaire in order to facilitate compliance and reduce the administrative burden for deployers.

+ (97) The notion of general-purpose AI models should be clearly defined and set apart from the notion of AI systems to enable legal certainty. The definition should be based on the key functional characteristics of a general-purpose AI model, in particular the generality and the capability to competently perform a wide range of distinct tasks. These models are typically trained on large amounts of data, through various methods, such as self-supervised, unsupervised or reinforcement learning. General-purpose AI models may be placed on the market in various ways, including through libraries, application programming interfaces (APIs), as direct download, or as physical copy. These models may be further modified or fine-tuned into new models. Although AI models are essential components of AI systems, they do not constitute AI systems on their own. AI models require the addition of further components, such as for example a user interface, to become AI systems. AI models are typically integrated into and form part of AI systems. This Regulation provides specific rules for general-purpose AI models and for general-purpose AI models that pose systemic risks, which should apply also when these models are integrated or form part of an AI system. It should be understood that the obligations for the providers of general-purpose AI models should apply once the general-purpose AI models are placed on the market. When the provider of a general-purpose AI model integrates an own model into its own AI system that is made available on the market or put into service, that model should be considered to be placed on the market and, therefore, the obligations in this Regulation for models should continue to apply in addition to those for AI systems. The obligations laid down for models should in any case not apply when an own model is used for purely internal processes that are not essential for providing a product or a service to third parties and the rights of natural persons are not affected. Considering their potential significantly negative effects, the general-purpose AI models with systemic risk should always be subject to the relevant obligations under this Regulation. The definition should not cover AI models used before their placing on the market for the sole purpose of research, development and prototyping activities. This is without prejudice to the obligation to comply with this Regulation when, following such activities, a model is placed on the market.

+ (98) Whereas the generality of a model could, inter alia, also be determined by a number of parameters, models with at least a billion of parameters and trained with a large amount of data using self-supervision at scale should be considered to display significant generality and to competently perform a wide range of distinctive tasks.

+ (99) Large generative AI models are a typical example for a general-purpose AI model, given that they allow for flexible generation of content, such as in the form of text, audio, images or video, that can readily accommodate a wide range of distinctive tasks.

+ (100) When a general-purpose AI model is integrated into or forms part of an AI system, this system should be considered to be general-purpose AI system when, due to this integration, this system has the capability to serve a variety of purposes. A general-purpose AI system can be used directly, or it may be integrated into other AI systems.

+ 26/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ (101) Providers of general-purpose AI models have a particular role and responsibility along the AI value chain, as the models they provide may form the basis for a range of downstream systems, often provided by downstream providers that necessitate a good understanding of the models and their capabilities, both to enable the integration of such models into their products, and to fulfil their obligations under this or other regulations. Therefore, proportionate transparency measures should be laid down, including the drawing up and keeping up to date of documentation, and the provision of information on the general-purpose AI model for its usage by the downstream providers. Technical documentation should be prepared and kept up to date by the general-purpose AI model provider for the purpose of making it available, upon request, to the AI Office and the national competent authorities. The minimal set of elements to be included in such documentation should be set out in specific annexes to this Regulation. The Commission should be empowered to amend those annexes by means of delegated acts in light of evolving technological developments.

+ (102) Software and data, including models, released under a free and open-source licence that allows them to be openly shared and where users can freely access, use, modify and redistribute them or modified versions thereof, can contribute to research and innovation in the market and can provide significant growth opportunities for the Union economy. General-purpose AI models released under free and open-source licences should be considered to ensure high levels of transparency and openness if their parameters, including the weights, the information on the model architecture, and the information on model usage are made publicly available. The licence should be considered to be free and open-source also when it allows users to run, copy, distribute, study, change and improve software and data, including models under the condition that the original provider of the model is credited, the identical or comparable terms of distribution are respected.

+ (103) Free and open-source AI components covers the software and data, including models and general-purpose AI models, tools, services or processes of an AI system. Free and open-source AI components can be provided through different channels, including their development on open repositories. For the purposes of this Regulation, AI components that are provided against a price or otherwise monetised, including through the provision of technical support or other services, including through a software platform, related to the AI component, or the use of personal data for reasons other than exclusively for improving the security, compatibility or interoperability of the software, with the exception of transactions between microenterprises, should not benefit from the exceptions provided to free and open-source AI components. The fact of making AI components available through open repositories should not, in itself, constitute a monetisation.

+ (104) The providers of general-purpose AI models that are released under a free and open-source licence, and whose parameters, including the weights, the information on the model architecture, and the information on model usage, are made publicly available should be subject to exceptions as regards the transparency-related requirements imposed on general-purpose AI models, unless they can be considered to present a systemic risk, in which case the circumstance that the model is transparent and accompanied by an open-source license should not be considered to be a sufficient reason to exclude compliance with the obligations under this Regulation. In any case, given that the release of general-purpose AI models under free and open-source licence does not necessarily reveal substantial information on the data set used for the training or fine-tuning of the model and on how compliance of copyright law was thereby ensured, the exception provided for general-purpose AI models from compliance with the transparency-related requirements should not concern the obligation to produce a summary about the content used for model training and the obligation to put in place a policy to comply with Union copyright law, in particular to identify and comply with the reservation of rights pursuant to Article 4(3) of Directive (EU) 2019/790 of the European Parliament and of the Council (40).

+ (105) General-purpose AI models, in particular large generative AI models, capable of generating text, images, and other content, present unique innovation opportunities but also challenges to artists, authors, and other creators and the way their creative content is created, distributed, used and consumed. The development and training of such models require access to vast amounts of text, images, videos and other data. Text and data mining techniques may be used extensively in this context for the retrieval and analysis of such content, which may be protected by copyright and related rights. Any use of copyright protected content requires the authorisation of the rightsholder concerned unless relevant copyright exceptions and limitations apply. Directive (EU) 2019/790 introduced exceptions and limitations allowing reproductions and extractions of works or other subject matter, for the purpose of text and data

+ (40) Directive (EU) 2019/790 of the European Parliament and of the Council of 17 April 2019 on copyright and related rights in the Digital Single Market and amending Directives 96/9/EC and 2001/29/EC (OJ L 130, 17.5.2019, p. 92).

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 27/144

+ mining, under certain conditions. Under these rules, rightsholders may choose to reserve their rights over their works or other subject matter to prevent text and data mining, unless this is done for the purposes of scientific research. Where the rights to opt out has been expressly reserved in an appropriate manner, providers of general-purpose AI models need to obtain an authorisation from rightsholders if they want to carry out text and data mining over such works.

+ (106) Providers that place general-purpose AI models on the Union market should ensure compliance with the relevant obligations in this Regulation. To that end, providers of general-purpose AI models should put in place a policy to comply with Union law on copyright and related rights, in particular to identify and comply with the reservation of rights expressed by rightsholders pursuant to Article 4(3) of Directive (EU) 2019/790. Any provider placing a general-purpose AI model on the Union market should comply with this obligation, regardless of the jurisdiction in which the copyright-relevant acts underpinning the training of those general-purpose AI models take place. This is necessary to ensure a level playing field among providers of general-purpose AI models where no provider should be able to gain a competitive advantage in the Union market by applying lower copyright standards than those provided in the Union.

+ (107) In order to increase transparency on the data that is used in the pre-training and training of general-purpose AI models, including text and data protected by copyright law, it is adequate that providers of such models draw up and make publicly available a sufficiently detailed summary of the content used for training the general-purpose AI model. While taking into due account the need to protect trade secrets and confidential business information, this summary should be generally comprehensive in its scope instead of technically detailed to facilitate parties with legitimate interests, including copyright holders, to exercise and enforce their rights under Union law, for example by listing the main data collections or sets that went into training the model, such as large private or public databases or data archives, and by providing a narrative explanation about other data sources used. It is appropriate for the AI Office to provide a template for the summary, which should be simple, effective, and allow the provider to provide the required summary in narrative form.

+ (108) With regard to the obligations imposed on providers of general-purpose AI models to put in place a policy to comply with Union copyright law and make publicly available a summary of the content used for the training, the AI Office should monitor whether the provider has fulfilled those obligations without verifying or proceeding to a work-by-work assessment of the training data in terms of copyright compliance. This Regulation does not affect the enforcement of copyright rules as provided for under Union law.

+ (109) Compliance with the obligations applicable to the providers of general-purpose AI models should be commensurate and proportionate to the type of model provider, excluding the need for compliance for persons who develop or use models for non-professional or scientific research purposes, who should nevertheless be encouraged to voluntarily comply with these requirements. Without prejudice to Union copyright law, compliance with those obligations should take due account of the size of the provider and allow simplified ways of compliance for SMEs, including start-ups, that should not represent an excessive cost and not discourage the use of such models. In the case of a modification or fine-tuning of a model, the obligations for providers of general-purpose AI models should be limited to that modification or fine-tuning, for example by complementing the already existing technical documentation with information on the modifications, including new training data sources, as a means to comply with the value chain obligations provided in this Regulation.

+ (110) General-purpose AI models could pose systemic risks which include, but are not limited to, any actual or reasonably foreseeable negative effects in relation to major accidents, disruptions of critical sectors and serious consequences to public health and safety; any actual or reasonably foreseeable negative effects on democratic processes, public and economic security; the dissemination of illegal, false, or discriminatory content. Systemic risks should be understood to increase with model capabilities and model reach, can arise along the entire lifecycle of the model, and are influenced by conditions of misuse, model reliability, model fairness and model security, the level of autonomy of

+ 28/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ the model, its access to tools, novel or combined modalities, release and distribution strategies, the potential to remove guardrails and other factors. In particular, international approaches have so far identified the need to pay attention to risks from potential intentional misuse or unintended issues of control relating to alignment with human intent; chemical, biological, radiological, and nuclear risks, such as the ways in which barriers to entry can be lowered, including for weapons development, design acquisition, or use; offensive cyber capabilities, such as the ways in vulnerability discovery, exploitation, or operational use can be enabled; the effects of interaction and tool use, including for example the capacity to control physical systems and interfere with critical infrastructure; risks from models of making copies of themselves or `self-replicating' or training other models; the ways in which models can give rise to harmful bias and discrimination with risks to individuals, communities or societies; the facilitation of disinformation or harming privacy with threats to democratic values and human rights; risk that a particular event could lead to a chain reaction with considerable negative effects that could affect up to an entire city, an entire domain activity or an entire community.

+ (111) It is appropriate to establish a methodology for the classification of general-purpose AI models as general-purpose AI model with systemic risks. Since systemic risks result from particularly high capabilities, a general-purpose AI model should be considered to present systemic risks if it has high-impact capabilities, evaluated on the basis of appropriate technical tools and methodologies, or significant impact on the internal market due to its reach. High-impact capabilities in general-purpose AI models means capabilities that match or exceed the capabilities recorded in the most advanced general-purpose AI models. The full range of capabilities in a model could be better understood after its placing on the market or when deployers interact with the model. According to the state of the art at the time of entry into force of this Regulation, the cumulative amount of computation used for the training of the general-purpose AI model measured in floating point operations is one of the relevant approximations for model capabilities. The cumulative amount of computation used for training includes the computation used across the activities and methods that are intended to enhance the capabilities of the model prior to deployment, such as pre-training, synthetic data generation and fine-tuning. Therefore, an initial threshold of floating point operations should be set, which, if met by a general-purpose AI model, leads to a presumption that the model is a general-purpose AI model with systemic risks. This threshold should be adjusted over time to reflect technological and industrial changes, such as algorithmic improvements or increased hardware efficiency, and should be supplemented with benchmarks and indicators for model capability. To inform this, the AI Office should engage with the scientific community, industry, civil society and other experts. Thresholds, as well as tools and benchmarks for the assessment of high-impact capabilities, should be strong predictors of generality, its capabilities and associated systemic risk of general-purpose AI models, and could take into account the way the model will be placed on the market or the number of users it may affect. To complement this system, there should be a possibility for the Commission to take individual decisions designating a general-purpose AI model as a general-purpose AI model with systemic risk if it is found that such model has capabilities or an impact equivalent to those captured by the set threshold. That decision should be taken on the basis of an overall assessment of the criteria for the designation of a general-purpose AI model with systemic risk set out in an annex to this Regulation, such as quality or size of the training data set, number of business and end users, its input and output modalities, its level of autonomy and scalability, or the tools it has access to. Upon a reasoned request of a provider whose model has been designated as a general-purpose AI model with systemic risk, the Commission should take the request into account and may decide to reassess whether the general-purpose AI model can still be considered to present systemic risks.

+ (112) It is also necessary to clarify a procedure for the classification of a general-purpose AI model with systemic risks. A general-purpose AI model that meets the applicable threshold for high-impact capabilities should be presumed to be a general-purpose AI models with systemic risk. The provider should notify the AI Office at the latest two weeks after the requirements are met or it becomes known that a general-purpose AI model will meet the requirements that lead to the presumption. This is especially relevant in relation to the threshold of floating point operations because training of general-purpose AI models takes considerable planning which includes the upfront allocation of compute resources and, therefore, providers of general-purpose AI models are able to know if their model would meet the threshold before the training is completed. In the context of that notification, the provider should be able to demonstrate that, because of its specific characteristics, a general-purpose AI model exceptionally does not present systemic risks, and that it thus should not be classified as a general-purpose AI model with systemic risks. That information is valuable for the AI Office to anticipate the placing on the market of general-purpose AI models with systemic risks and the providers can start to engage with the AI Office early on. That information is especially

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 29/144

+ important with regard to general-purpose AI models that are planned to be released as open-source, given that, after the open-source model release, necessary measures to ensure compliance with the obligations under this Regulation may be more difficult to implement.

+ (113) If the Commission becomes aware of the fact that a general-purpose AI model meets the requirements to classify as a general-purpose AI model with systemic risk, which previously had either not been known or of which the relevant provider has failed to notify the Commission, the Commission should be empowered to designate it so. A system of qualified alerts should ensure that the AI Office is made aware by the scientific panel of general-purpose AI models that should possibly be classified as general-purpose AI models with systemic risk, in addition to the monitoring activities of the AI Office.

+ (114) The providers of general-purpose AI models presenting systemic risks should be subject, in addition to the obligations provided for providers of general-purpose AI models, to obligations aimed at identifying and mitigating those risks and ensuring an adequate level of cybersecurity protection, regardless of whether it is provided as a standalone model or embedded in an AI system or a product. To achieve those objectives, this Regulation should require providers to perform the necessary model evaluations, in particular prior to its first placing on the market, including conducting and documenting adversarial testing of models, also, as appropriate, through internal or independent external testing. In addition, providers of general-purpose AI models with systemic risks should continuously assess and mitigate systemic risks, including for example by putting in place risk-management policies, such as accountability and governance processes, implementing post-market monitoring, taking appropriate measures along the entire model's lifecycle and cooperating with relevant actors along the AI value chain.

+ (115) Providers of general-purpose AI models with systemic risks should assess and mitigate possible systemic risks. If, despite efforts to identify and prevent risks related to a general-purpose AI model that may present systemic risks, the development or use of the model causes a serious incident, the general-purpose AI model provider should without undue delay keep track of the incident and report any relevant information and possible corrective measures to the Commission and national competent authorities. Furthermore, providers should ensure an adequate level of cybersecurity protection for the model and its physical infrastructure, if appropriate, along the entire model lifecycle. Cybersecurity protection related to systemic risks associated with malicious use or attacks should duly consider accidental model leakage, unauthorised releases, circumvention of safety measures, and defence against cyberattacks, unauthorised access or model theft. That protection could be facilitated by securing model weights, algorithms, servers, and data sets, such as through operational security measures for information security, specific cybersecurity policies, adequate technical and established solutions, and cyber and physical access controls, appropriate to the relevant circumstances and the risks involved.

+ (116) The AI Office should encourage and facilitate the drawing up, review and adaptation of codes of practice, taking into account international approaches. All providers of general-purpose AI models could be invited to participate. To ensure that the codes of practice reflect the state of the art and duly take into account a diverse set of perspectives, the AI Office should collaborate with relevant national competent authorities, and could, where appropriate, consult with civil society organisations and other relevant stakeholders and experts, including the Scientific Panel, for the drawing up of such codes. Codes of practice should cover obligations for providers of general-purpose AI models and of general-purpose AI models presenting systemic risks. In addition, as regards systemic risks, codes of practice should help to establish a risk taxonomy of the type and nature of the systemic risks at Union level, including their sources. Codes of practice should also be focused on specific risk assessment and mitigation measures.

+ (117) The codes of practice should represent a central tool for the proper compliance with the obligations provided for under this Regulation for providers of general-purpose AI models. Providers should be able to rely on codes of practice to demonstrate compliance with the obligations. By means of implementing acts, the Commission may decide to approve a code of practice and give it a general validity within the Union, or, alternatively, to provide common rules for the implementation of the relevant obligations, if, by the time this Regulation becomes applicable, a code of practice cannot be finalised or is not deemed adequate by the AI Office. Once a harmonised standard is

+ 30/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ published and assessed as suitable to cover the relevant obligations by the AI Office, compliance with a European harmonised standard should grant providers the presumption of conformity. Providers of general-purpose AI models should furthermore be able to demonstrate compliance using alternative adequate means, if codes of practice or harmonised standards are not available, or they choose not to rely on those.

+ (118) This Regulation regulates AI systems and AI models by imposing certain requirements and obligations for relevant market actors that are placing them on the market, putting into service or use in the Union, thereby complementing obligations for providers of intermediary services that embed such systems or models into their services regulated by Regulation (EU) 2022/2065. To the extent that such systems or models are embedded into designated very large online platforms or very large online search engines, they are subject to the risk-management framework provided for in Regulation (EU) 2022/2065. Consequently, the corresponding obligations of this Regulation should be presumed to be fulfilled, unless significant systemic risks not covered by Regulation (EU) 2022/2065 emerge and are identified in such models. Within this framework, providers of very large online platforms and very large online search engines are obliged to assess potential systemic risks stemming from the design, functioning and use of their services, including how the design of algorithmic systems used in the service may contribute to such risks, as well as systemic risks stemming from potential misuses. Those providers are also obliged to take appropriate mitigating measures in observance of fundamental rights.

+ (119) Considering the quick pace of innovation and the technological evolution of digital services in scope of different instruments of Union law in particular having in mind the usage and the perception of their recipients, the AI systems subject to this Regulation may be provided as intermediary services or parts thereof within the meaning of Regulation (EU) 2022/2065, which should be interpreted in a technology-neutral manner. For example, AI systems may be used to provide online search engines, in particular, to the extent that an AI system such as an online chatbot performs searches of, in principle, all websites, then incorporates the results into its existing knowledge and uses the updated knowledge to generate a single output that combines different sources of information.

+ (120) Furthermore, obligations placed on providers and deployers of certain AI systems in this Regulation to enable the detection and disclosure that the outputs of those systems are artificially generated or manipulated are particularly relevant to facilitate the effective implementation of Regulation (EU) 2022/2065. This applies in particular as regards the obligations of providers of very large online platforms or very large online search engines to identify and mitigate systemic risks that may arise from the dissemination of content that has been artificially generated or manipulated, in particular risk of the actual or foreseeable negative effects on democratic processes, civic discourse and electoral processes, including through disinformation.

+ (121) Standardisation should play a key role to provide technical solutions to providers to ensure compliance with this Regulation, in line with the state of the art, to promote innovation as well as competitiveness and growth in the single market. Compliance with harmonised standards as defined in Article 2, point (1)(c), of Regulation (EU) No 1025/2012 of the European Parliament and of the Council (41), which are normally expected to reflect the state of the art, should be a means for providers to demonstrate conformity with the requirements of this Regulation. A balanced representation of interests involving all relevant stakeholders in the development of standards, in particular SMEs, consumer organisations and environmental and social stakeholders in accordance with Articles 5 and 6 of Regulation (EU) No 1025/2012 should therefore be encouraged. In order to facilitate compliance, the standardisation requests should be issued by the Commission without undue delay. When preparing the standardisation request, the Commission should consult the advisory forum and the Board in order to collect relevant expertise. However, in the absence of relevant references to harmonised standards, the Commission should be able to establish, via implementing acts, and after consultation of the advisory forum, common specifications for certain requirements under this Regulation. The common specification should be an exceptional fall back solution to facilitate the provider's obligation to comply with the requirements of this Regulation, when the standardisation request has not been accepted by any of the European standardisation organisations, or when the relevant harmonised standards insufficiently address fundamental rights concerns, or when the harmonised standards do not comply with the request, or when there are delays in the adoption of an appropriate harmonised standard. Where such a delay in the adoption of a harmonised standard is due to the technical complexity of that standard, this should

+ (41) Regulation (EU) No 1025/2012 of the European Parliament and of the Council of 25 October 2012 on European standardisation, amending Council Directives 89/686/EEC and 93/15/EEC and Directives 94/9/EC, 94/25/EC, 95/16/EC, 97/23/EC, 98/34/EC, 2004/22/EC, 2007/23/EC, 2009/23/EC and 2009/105/EC of the European Parliament and of the Council and repealing Council Decision 87/95/EEC and Decision No 1673/2006/EC of the European Parliament and of the Council (OJ L 316, 14.11.2012, p. 12).

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 31/144

+ be considered by the Commission before contemplating the establishment of common specifications. When developing common specifications, the Commission is encouraged to cooperate with international partners and international standardisation bodies.

+ (122) It is appropriate that, without prejudice to the use of harmonised standards and common specifications, providers of a high-risk AI system that has been trained and tested on data reflecting the specific geographical, behavioural, contextual or functional setting within which the AI system is intended to be used, should be presumed to comply with the relevant measure provided for under the requirement on data governance set out in this Regulation. Without prejudice to the requirements related to robustness and accuracy set out in this Regulation, in accordance with Article 54(3) of Regulation (EU) 2019/881, high-risk AI systems that have been certified or for which a statement of conformity has been issued under a cybersecurity scheme pursuant to that Regulation and the references of which have been published in the Official Journal of the European Union should be presumed to comply with the cybersecurity requirement of this Regulation in so far as the cybersecurity certificate or statement of conformity or parts thereof cover the cybersecurity requirement of this Regulation. This remains without prejudice to the voluntary nature of that cybersecurity scheme.

+ (123) In order to ensure a high level of trustworthiness of high-risk AI systems, those systems should be subject to a conformity assessment prior to their placing on the market or putting into service.

+ (124) It is appropriate that, in order to minimise the burden on operators and avoid any possible duplication, for high-risk AI systems related to products which are covered by existing Union harmonisation legislation based on the New Legislative Framework, the compliance of those AI systems with the requirements of this Regulation should be assessed as part of the conformity assessment already provided for in that law. The applicability of the requirements of this Regulation should thus not affect the specific logic, methodology or general structure of conformity assessment under the relevant Union harmonisation legislation.

+ (125) Given the complexity of high-risk AI systems and the risks that are associated with them, it is important to develop an adequate conformity assessment procedure for high-risk AI systems involving notified bodies, so-called third party conformity assessment. However, given the current experience of professional pre-market certifiers in the field of product safety and the different nature of risks involved, it is appropriate to limit, at least in an initial phase of application of this Regulation, the scope of application of third-party conformity assessment for high-risk AI systems other than those related to products. Therefore, the conformity assessment of such systems should be carried out as a general rule by the provider under its own responsibility, with the only exception of AI systems intended to be used for biometrics.

+ (126) In order to carry out third-party conformity assessments when so required, notified bodies should be notified under this Regulation by the national competent authorities, provided that they comply with a set of requirements, in particular on independence, competence, absence of conflicts of interests and suitable cybersecurity requirements. Notification of those bodies should be sent by national competent authorities to the Commission and the other Member States by means of the electronic notification tool developed and managed by the Commission pursuant to Article R23 of Annex I to Decision No 768/2008/EC.

+ (127) In line with Union commitments under the World Trade Organization Agreement on Technical Barriers to Trade, it is adequate to facilitate the mutual recognition of conformity assessment results produced by competent conformity assessment bodies, independent of the territory in which they are established, provided that those conformity assessment bodies established under the law of a third country meet the applicable requirements of this Regulation and the Union has concluded an agreement to that extent. In this context, the Commission should actively explore possible international instruments for that purpose and in particular pursue the conclusion of mutual recognition agreements with third countries.

+ (128) In line with the commonly established notion of substantial modification for products regulated by Union harmonisation legislation, it is appropriate that whenever a change occurs which may affect the compliance of a high-risk AI system with this Regulation (e.g. change of operating system or software architecture), or when the intended purpose of the system changes, that AI system should be considered to be a new AI system which should undergo a new conformity assessment. However, changes occurring to the algorithm and the performance of AI systems which continue to `learn' after being placed on the market or put into service, namely automatically adapting how functions are carried out, should not constitute a substantial modification, provided that those changes have been pre-determined by the provider and assessed at the moment of the conformity assessment.

+ 32/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ (129) High-risk AI systems should bear the CE marking to indicate their conformity with this Regulation so that they can move freely within the internal market. For high-risk AI systems embedded in a product, a physical CE marking should be affixed, and may be complemented by a digital CE marking. For high-risk AI systems only provided digitally, a digital CE marking should be used. Member States should not create unjustified obstacles to the placing on the market or the putting into service of high-risk AI systems that comply with the requirements laid down in this Regulation and bear the CE marking.

+ (130) Under certain conditions, rapid availability of innovative technologies may be crucial for health and safety of persons, the protection of the environment and climate change and for society as a whole. It is thus appropriate that under exceptional reasons of public security or protection of life and health of natural persons, environmental protection and the protection of key industrial and infrastructural assets, market surveillance authorities could authorise the placing on the market or the putting into service of AI systems which have not undergone a conformity assessment. In duly justified situations, as provided for in this Regulation, law enforcement authorities or civil protection authorities may put a specific high-risk AI system into service without the authorisation of the market surveillance authority, provided that such authorisation is requested during or after the use without undue delay.

+ (131) In order to facilitate the work of the Commission and the Member States in the AI field as well as to increase the transparency towards the public, providers of high-risk AI systems other than those related to products falling within the scope of relevant existing Union harmonisation legislation, as well as providers who consider that an AI system listed in the high-risk use cases in an annex to this Regulation is not high-risk on the basis of a derogation, should be required to register themselves and information about their AI system in an EU database, to be established and managed by the Commission. Before using an AI system listed in the high-risk use cases in an annex to this Regulation, deployers of high-risk AI systems that are public authorities, agencies or bodies, should register themselves in such database and select the system that they envisage to use. Other deployers should be entitled to do so voluntarily. This section of the EU database should be publicly accessible, free of charge, the information should be easily navigable, understandable and machine-readable. The EU database should also be user-friendly, for example by providing search functionalities, including through keywords, allowing the general public to find relevant information to be submitted upon the registration of high-risk AI systems and on the use case of high-risk AI systems, set out in an annex to this Regulation, to which the high-risk AI systems correspond. Any substantial modification of high-risk AI systems should also be registered in the EU database. For high-risk AI systems in the area of law enforcement, migration, asylum and border control management, the registration obligations should be fulfilled in a secure non-public section of the EU database. Access to the secure non-public section should be strictly limited to the Commission as well as to market surveillance authorities with regard to their national section of that database. High-risk AI systems in the area of critical infrastructure should only be registered at national level. The Commission should be the controller of the EU database, in accordance with Regulation (EU) 2018/1725. In order to ensure the full functionality of the EU database, when deployed, the procedure for setting the database should include the development of functional specifications by the Commission and an independent audit report. The Commission should take into account cybersecurity risks when carrying out its tasks as data controller on the EU database. In order to maximise the availability and use of the EU database by the public, the EU database, including the information made available through it, should comply with requirements under the Directive (EU) 2019/882.

+ (132) Certain AI systems intended to interact with natural persons or to generate content may pose specific risks of impersonation or deception irrespective of whether they qualify as high-risk or not. In certain circumstances, the use of these systems should therefore be subject to specific transparency obligations without prejudice to the requirements and obligations for high-risk AI systems and subject to targeted exceptions to take into account the special need of law enforcement. In particular, natural persons should be notified that they are interacting with an AI system, unless this is obvious from the point of view of a natural person who is reasonably well-informed, observant and circumspect taking into account the circumstances and the context of use. When implementing that obligation, the characteristics of natural persons belonging to vulnerable groups due to their age or disability should be taken into account to the extent the AI system is intended to interact with those groups as well. Moreover, natural persons should be notified when they are exposed to AI systems that, by processing their biometric data, can identify or infer the emotions or intentions of those persons or assign them to specific categories. Such specific categories can relate to aspects such as sex, age, hair colour, eye colour, tattoos, personal traits, ethnic origin, personal preferences and interests. Such information and notifications should be provided in accessible formats for persons with disabilities.

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 33/144

+ (133) A variety of AI systems can generate large quantities of synthetic content that becomes increasingly hard for humans to distinguish from human-generated and authentic content. The wide availability and increasing capabilities of those systems have a significant impact on the integrity and trust in the information ecosystem, raising new risks of misinformation and manipulation at scale, fraud, impersonation and consumer deception. In light of those impacts, the fast technological pace and the need for new methods and techniques to trace origin of information, it is appropriate to require providers of those systems to embed technical solutions that enable marking in a machine readable format and detection that the output has been generated or manipulated by an AI system and not a human. Such techniques and methods should be sufficiently reliable, interoperable, effective and robust as far as this is technically feasible, taking into account available techniques or a combination of such techniques, such as watermarks, metadata identifications, cryptographic methods for proving provenance and authenticity of content, logging methods, fingerprints or other techniques, as may be appropriate. When implementing this obligation, providers should also take into account the specificities and the limitations of the different types of content and the relevant technological and market developments in the field, as reflected in the generally acknowledged state of the art. Such techniques and methods can be implemented at the level of the AI system or at the level of the AI model, including general-purpose AI models generating content, thereby facilitating fulfilment of this obligation by the downstream provider of the AI system. To remain proportionate, it is appropriate to envisage that this marking obligation should not cover AI systems performing primarily an assistive function for standard editing or AI systems not substantially altering the input data provided by the deployer or the semantics thereof.

+ (134) Further to the technical solutions employed by the providers of the AI system, deployers who use an AI system to generate or manipulate image, audio or video content that appreciably resembles existing persons, objects, places, entities or events and would falsely appear to a person to be authentic or truthful (deep fakes), should also clearly and distinguishably disclose that the content has been artificially created or manipulated by labelling the AI output accordingly and disclosing its artificial origin. Compliance with this transparency obligation should not be interpreted as indicating that the use of the AI system or its output impedes the right to freedom of expression and the right to freedom of the arts and sciences guaranteed in the Charter, in particular where the content is part of an evidently creative, satirical, artistic, fictional or analogous work or programme, subject to appropriate safeguards for the rights and freedoms of third parties. In those cases, the transparency obligation for deep fakes set out in this Regulation is limited to disclosure of the existence of such generated or manipulated content in an appropriate manner that does not hamper the display or enjoyment of the work, including its normal exploitation and use, while maintaining the utility and quality of the work. In addition, it is also appropriate to envisage a similar disclosure obligation in relation to AI-generated or manipulated text to the extent it is published with the purpose of informing the public on matters of public interest unless the AI-generated content has undergone a process of human review or editorial control and a natural or legal person holds editorial responsibility for the publication of the content.

+ (135) Without prejudice to the mandatory nature and full applicability of the transparency obligations, the Commission may also encourage and facilitate the drawing up of codes of practice at Union level to facilitate the effective implementation of the obligations regarding the detection and labelling of artificially generated or manipulated content, including to support practical arrangements for making, as appropriate, the detection mechanisms accessible and facilitating cooperation with other actors along the value chain, disseminating content or checking its authenticity and provenance to enable the public to effectively distinguish AI-generated content.

+ (136) The obligations placed on providers and deployers of certain AI systems in this Regulation to enable the detection and disclosure that the outputs of those systems are artificially generated or manipulated are particularly relevant to facilitate the effective implementation of Regulation (EU) 2022/2065. This applies in particular as regards the obligations of providers of very large online platforms or very large online search engines to identify and mitigate systemic risks that may arise from the dissemination of content that has been artificially generated or manipulated, in particular the risk of the actual or foreseeable negative effects on democratic processes, civic discourse and electoral processes, including through disinformation. The requirement to label content generated by AI systems under this Regulation is without prejudice to the obligation in Article 16(6) of Regulation (EU) 2022/2065 for providers of hosting services to process notices on illegal content received pursuant to Article 16(1) of that Regulation and should not influence the assessment and the decision on the illegality of the specific content. That assessment should be performed solely with reference to the rules governing the legality of the content.

+ 34/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ (137) Compliance with the transparency obligations for the AI systems covered by this Regulation should not be interpreted as indicating that the use of the AI system or its output is lawful under this Regulation or other Union and Member State law and should be without prejudice to other transparency obligations for deployers of AI systems laid down in Union or national law.

+ (138) AI is a rapidly developing family of technologies that requires regulatory oversight and a safe and controlled space for experimentation, while ensuring responsible innovation and integration of appropriate safeguards and risk mitigation measures. To ensure a legal framework that promotes innovation, is future-proof and resilient to disruption, Member States should ensure that their national competent authorities establish at least one AI regulatory sandbox at national level to facilitate the development and testing of innovative AI systems under strict regulatory oversight before these systems are placed on the market or otherwise put into service. Member States could also fulfil this obligation through participating in already existing regulatory sandboxes or establishing jointly a sandbox with one or more Member States' competent authorities, insofar as this participation provides equivalent level of national coverage for the participating Member States. AI regulatory sandboxes could be established in physical, digital or hybrid form and may accommodate physical as well as digital products. Establishing authorities should also ensure that the AI regulatory sandboxes have the adequate resources for their functioning, including financial and human resources.

+ (139) The objectives of the AI regulatory sandboxes should be to foster AI innovation by establishing a controlled experimentation and testing environment in the development and pre-marketing phase with a view to ensuring compliance of the innovative AI systems with this Regulation and other relevant Union and national law. Moreover, the AI regulatory sandboxes should aim to enhance legal certainty for innovators and the competent authorities' oversight and understanding of the opportunities, emerging risks and the impacts of AI use, to facilitate regulatory learning for authorities and undertakings, including with a view to future adaptions of the legal framework, to support cooperation and the sharing of best practices with the authorities involved in the AI regulatory sandbox, and to accelerate access to markets, including by removing barriers for SMEs, including start-ups. AI regulatory sandboxes should be widely available throughout the Union, and particular attention should be given to their accessibility for SMEs, including start-ups. The participation in the AI regulatory sandbox should focus on issues that raise legal uncertainty for providers and prospective providers to innovate, experiment with AI in the Union and contribute to evidence-based regulatory learning. The supervision of the AI systems in the AI regulatory sandbox should therefore cover their development, training, testing and validation before the systems are placed on the market or put into service, as well as the notion and occurrence of substantial modification that may require a new conformity assessment procedure. Any significant risks identified during the development and testing of such AI systems should result in adequate mitigation and, failing that, in the suspension of the development and testing process. Where appropriate, national competent authorities establishing AI regulatory sandboxes should cooperate with other relevant authorities, including those supervising the protection of fundamental rights, and could allow for the involvement of other actors within the AI ecosystem such as national or European standardisation organisations, notified bodies, testing and experimentation facilities, research and experimentation labs, European Digital Innovation Hubs and relevant stakeholder and civil society organisations. To ensure uniform implementation across the Union and economies of scale, it is appropriate to establish common rules for the AI regulatory sandboxes' implementation and a framework for cooperation between the relevant authorities involved in the supervision of the sandboxes. AI regulatory sandboxes established under this Regulation should be without prejudice to other law allowing for the establishment of other sandboxes aiming to ensure compliance with law other than this Regulation. Where appropriate, relevant competent authorities in charge of those other regulatory sandboxes should consider the benefits of using those sandboxes also for the purpose of ensuring compliance of AI systems with this Regulation. Upon agreement between the national competent authorities and the participants in the AI regulatory sandbox, testing in real world conditions may also be operated and supervised in the framework of the AI regulatory sandbox.

+ (140) This Regulation should provide the legal basis for the providers and prospective providers in the AI regulatory sandbox to use personal data collected for other purposes for developing certain AI systems in the public interest within the AI regulatory sandbox, only under specified conditions, in accordance with Article 6(4) and Article 9(2), point (g), of Regulation (EU) 2016/679, and Articles 5, 6 and 10 of Regulation (EU) 2018/1725, and without prejudice to Article 4(2) and Article 10 of Directive (EU) 2016/680. All other obligations of data controllers and rights of data subjects under Regulations (EU) 2016/679 and (EU) 2018/1725 and Directive (EU) 2016/680 remain applicable. In particular, this Regulation should not provide a legal basis in the meaning of Article 22(2), point (b) of Regulation (EU) 2016/679 and Article 24(2), point (b) of Regulation (EU) 2018/1725. Providers and prospective

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 35/144

+ providers in the AI regulatory sandbox should ensure appropriate safeguards and cooperate with the competent authorities, including by following their guidance and acting expeditiously and in good faith to adequately mitigate any identified significant risks to safety, health, and fundamental rights that may arise during the development, testing and experimentation in that sandbox.

+ (141) In order to accelerate the process of development and the placing on the market of the high-risk AI systems listed in an annex to this Regulation, it is important that providers or prospective providers of such systems may also benefit from a specific regime for testing those systems in real world conditions, without participating in an AI regulatory sandbox. However, in such cases, taking into account the possible consequences of such testing on individuals, it should be ensured that appropriate and sufficient guarantees and conditions are introduced by this Regulation for providers or prospective providers. Such guarantees should include, inter alia, requesting informed consent of natural persons to participate in testing in real world conditions, with the exception of law enforcement where the seeking of informed consent would prevent the AI system from being tested. Consent of subjects to participate in such testing under this Regulation is distinct from, and without prejudice to, consent of data subjects for the processing of their personal data under the relevant data protection law. It is also important to minimise the risks and enable oversight by competent authorities and therefore require prospective providers to have a real-world testing plan submitted to competent market surveillance authority, register the testing in dedicated sections in the EU database subject to some limited exceptions, set limitations on the period for which the testing can be done and require additional safeguards for persons belonging to certain vulnerable groups, as well as a written agreement defining the roles and responsibilities of prospective providers and deployers and effective oversight by competent personnel involved in the real world testing. Furthermore, it is appropriate to envisage additional safeguards to ensure that the predictions, recommendations or decisions of the AI system can be effectively reversed and disregarded and that personal data is protected and is deleted when the subjects have withdrawn their consent to participate in the testing without prejudice to their rights as data subjects under the Union data protection law. As regards transfer of data, it is also appropriate to envisage that data collected and processed for the purpose of testing in real-world conditions should be transferred to third countries only where appropriate and applicable safeguards under Union law are implemented, in particular in accordance with bases for transfer of personal data under Union law on data protection, while for non-personal data appropriate safeguards are put in place in accordance with Union law, such as Regulations (EU) 2022/868 (42) and (EU) 2023/2854 (43) of the European Parliament and of the Council.

+ (142) To ensure that AI leads to socially and environmentally beneficial outcomes, Member States are encouraged to support and promote research and development of AI solutions in support of socially and environmentally beneficial outcomes, such as AI-based solutions to increase accessibility for persons with disabilities, tackle socio-economic inequalities, or meet environmental targets, by allocating sufficient resources, including public and Union funding, and, where appropriate and provided that the eligibility and selection criteria are fulfilled, considering in particular projects which pursue such objectives. Such projects should be based on the principle of interdisciplinary cooperation between AI developers, experts on inequality and non-discrimination, accessibility, consumer, environmental, and digital rights, as well as academics.

+ (143) In order to promote and protect innovation, it is important that the interests of SMEs, including start-ups, that are providers or deployers of AI systems are taken into particular account. To that end, Member States should develop initiatives, which are targeted at those operators, including on awareness raising and information communication. Member States should provide SMEs, including start-ups, that have a registered office or a branch in the Union, with priority access to the AI regulatory sandboxes provided that they fulfil the eligibility conditions and selection criteria and without precluding other providers and prospective providers to access the sandboxes provided the same conditions and criteria are fulfilled. Member States should utilise existing channels and where appropriate, establish new dedicated channels for communication with SMEs, including start-ups, deployers, other innovators and, as appropriate, local public authorities, to support SMEs throughout their development path by providing guidance and responding to queries about the implementation of this Regulation. Where appropriate, these channels should work together to create synergies and ensure homogeneity in their guidance to SMEs, including start-ups, and deployers. Additionally, Member States should facilitate the participation of SMEs and other relevant stakeholders in the standardisation development processes. Moreover, the specific interests and needs of providers that are SMEs,

+ (42) Regulation (EU) 2022/868 of the European Parliament and of the Council of 30 May 2022 on European data governance and amending Regulation (EU) 2018/1724 (Data Governance Act) (OJ L 152, 3.6.2022, p. 1).

+ (43) Regulation (EU) 2023/2854 of the European Parliament and of the Council of 13 December 2023 on harmonised rules on fair access to and use of data and amending Regulation (EU) 2017/2394 and Directive (EU) 2020/1828 (Data Act) (OJ L, 2023/2854, 22.12.2023, ELI: http://data.europa.eu/eli/reg/2023/2854/oj).

+ 36/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ including start-ups, should be taken into account when notified bodies set conformity assessment fees. The Commission should regularly assess the certification and compliance costs for SMEs, including start-ups, through transparent consultations and should work with Member States to lower such costs. For example, translation costs related to mandatory documentation and communication with authorities may constitute a significant cost for providers and other operators, in particular those of a smaller scale. Member States should possibly ensure that one of the languages determined and accepted by them for relevant providers' documentation and for communication with operators is one which is broadly understood by the largest possible number of cross-border deployers. In order to address the specific needs of SMEs, including start-ups, the Commission should provide standardised templates for the areas covered by this Regulation, upon request of the Board. Additionally, the Commission should complement Member States' efforts by providing a single information platform with easy-to-use information with regards to this Regulation for all providers and deployers, by organising appropriate communication campaigns to raise awareness about the obligations arising from this Regulation, and by evaluating and promoting the convergence of best practices in public procurement procedures in relation to AI systems. Medium-sized enterprises which until recently qualified as small enterprises within the meaning of the Annex to Commission Recommendation 2003/361/EC (44) should have access to those support measures, as those new medium-sized enterprises may sometimes lack the legal resources and training necessary to ensure proper understanding of, and compliance with, this Regulation.

+ (144) In order to promote and protect innovation, the AI-on-demand platform, all relevant Union funding programmes and projects, such as Digital Europe Programme, Horizon Europe, implemented by the Commission and the Member States at Union or national level should, as appropriate, contribute to the achievement of the objectives of this Regulation.

+ (145) In order to minimise the risks to implementation resulting from lack of knowledge and expertise in the market as well as to facilitate compliance of providers, in particular SMEs, including start-ups, and notified bodies with their obligations under this Regulation, the AI-on-demand platform, the European Digital Innovation Hubs and the testing and experimentation facilities established by the Commission and the Member States at Union or national level should contribute to the implementation of this Regulation. Within their respective mission and fields of competence, the AI-on-demand platform, the European Digital Innovation Hubs and the testing and experimentation Facilities are able to provide in particular technical and scientific support to providers and notified bodies.

+ (146) Moreover, in light of the very small size of some operators and in order to ensure proportionality regarding costs of innovation, it is appropriate to allow microenterprises to fulfil one of the most costly obligations, namely to establish a quality management system, in a simplified manner which would reduce the administrative burden and the costs for those enterprises without affecting the level of protection and the need for compliance with the requirements for high-risk AI systems. The Commission should develop guidelines to specify the elements of the quality management system to be fulfilled in this simplified manner by microenterprises.

+ (147) It is appropriate that the Commission facilitates, to the extent possible, access to testing and experimentation facilities to bodies, groups or laboratories established or accredited pursuant to any relevant Union harmonisation legislation and which fulfil tasks in the context of conformity assessment of products or devices covered by that Union harmonisation legislation. This is, in particular, the case as regards expert panels, expert laboratories and reference laboratories in the field of medical devices pursuant to Regulations (EU) 2017/745 and (EU) 2017/746.

+ (148) This Regulation should establish a governance framework that both allows to coordinate and support the application of this Regulation at national level, as well as build capabilities at Union level and integrate stakeholders in the field of AI. The effective implementation and enforcement of this Regulation require a governance framework that allows to coordinate and build up central expertise at Union level. The AI Office was established by Commission Decision (45) and has as its mission to develop Union expertise and capabilities in the field of AI and to contribute to the implementation of Union law on AI. Member States should facilitate the tasks of the AI Office with a view to support the development of Union expertise and capabilities at Union level and to strengthen the functioning of the digital single market. Furthermore, a Board composed of representatives of the Member States, a scientific panel to integrate the scientific community and an advisory forum to contribute stakeholder input to the implementation of this Regulation, at Union and national level, should be established. The development of Union expertise and

+ (44) Commission Recommendation of 6 May 2003 concerning the definition of micro, small and medium-sized enterprises (OJ L 124, 20.5.2003, p. 36).

+ (45) Commission Decision of 24.1.2024 establishing the European Artificial Intelligence Office C(2024) 390.

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 37/144

+ capabilities should also include making use of existing resources and expertise, in particular through synergies with structures built up in the context of the Union level enforcement of other law and synergies with related initiatives at Union level, such as the EuroHPC Joint Undertaking and the AI testing and experimentation facilities under the Digital Europe Programme.

+ (149) In order to facilitate a smooth, effective and harmonised implementation of this Regulation a Board should be established. The Board should reflect the various interests of the AI eco-system and be composed of representatives of the Member States. The Board should be responsible for a number of advisory tasks, including issuing opinions, recommendations, advice or contributing to guidance on matters related to the implementation of this Regulation, including on enforcement matters, technical specifications or existing standards regarding the requirements established in this Regulation and providing advice to the Commission and the Member States and their national competent authorities on specific questions related to AI. In order to give some flexibility to Member States in the designation of their representatives in the Board, such representatives may be any persons belonging to public entities who should have the relevant competences and powers to facilitate coordination at national level and contribute to the achievement of the Board's tasks. The Board should establish two standing sub-groups to provide a platform for cooperation and exchange among market surveillance authorities and notifying authorities on issues related, respectively, to market surveillance and notified bodies. The standing subgroup for market surveillance should act as the administrative cooperation group (ADCO) for this Regulation within the meaning of Article 30 of Regulation (EU) 2019/1020. In accordance with Article 33 of that Regulation, the Commission should support the activities of the standing subgroup for market surveillance by undertaking market evaluations or studies, in particular with a view to identifying aspects of this Regulation requiring specific and urgent coordination among market surveillance authorities. The Board may establish other standing or temporary sub-groups as appropriate for the purpose of examining specific issues. The Board should also cooperate, as appropriate, with relevant Union bodies, experts groups and networks active in the context of relevant Union law, including in particular those active under relevant Union law on data, digital products and services.

+ (150) With a view to ensuring the involvement of stakeholders in the implementation and application of this Regulation, an advisory forum should be established to advise and provide technical expertise to the Board and the Commission. To ensure a varied and balanced stakeholder representation between commercial and non-commercial interest and, within the category of commercial interests, with regards to SMEs and other undertakings, the advisory forum should comprise inter alia industry, start-ups, SMEs, academia, civil society, including the social partners, as well as the Fundamental Rights Agency, ENISA, the European Committee for Standardization (CEN), the European Committee for Electrotechnical Standardization (CENELEC) and the European Telecommunications Standards Institute (ETSI).

+ (151) To support the implementation and enforcement of this Regulation, in particular the monitoring activities of the AI Office as regards general-purpose AI models, a scientific panel of independent experts should be established. The independent experts constituting the scientific panel should be selected on the basis of up-to-date scientific or technical expertise in the field of AI and should perform their tasks with impartiality, objectivity and ensure the confidentiality of information and data obtained in carrying out their tasks and activities. To allow the reinforcement of national capacities necessary for the effective enforcement of this Regulation, Member States should be able to request support from the pool of experts constituting the scientific panel for their enforcement activities.

+ (152) In order to support adequate enforcement as regards AI systems and reinforce the capacities of the Member States, Union AI testing support structures should be established and made available to the Member States.

+ (153) Member States hold a key role in the application and enforcement of this Regulation. In that respect, each Member State should designate at least one notifying authority and at least one market surveillance authority as national competent authorities for the purpose of supervising the application and implementation of this Regulation. Member States may decide to appoint any kind of public entity to perform the tasks of the national competent authorities within the meaning of this Regulation, in accordance with their specific national organisational characteristics and needs. In order to increase organisation efficiency on the side of Member States and to set a single point of contact vis-�-vis the public and other counterparts at Member State and Union levels, each Member State should designate a market surveillance authority to act as a single point of contact.

+ 38/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ (154) The national competent authorities should exercise their powers independently, impartially and without bias, so as to safeguard the principles of objectivity of their activities and tasks and to ensure the application and implementation of this Regulation. The members of these authorities should refrain from any action incompatible with their duties and should be subject to confidentiality rules under this Regulation.

+ (155) In order to ensure that providers of high-risk AI systems can take into account the experience on the use of high-risk AI systems for improving their systems and the design and development process or can take any possible corrective action in a timely manner, all providers should have a post-market monitoring system in place. Where relevant, post-market monitoring should include an analysis of the interaction with other AI systems including other devices and software. Post-market monitoring should not cover sensitive operational data of deployers which are law enforcement authorities. This system is also key to ensure that the possible risks emerging from AI systems which continue to `learn' after being placed on the market or put into service can be more efficiently and timely addressed. In this context, providers should also be required to have a system in place to report to the relevant authorities any serious incidents resulting from the use of their AI systems, meaning incident or malfunctioning leading to death or serious damage to health, serious and irreversible disruption of the management and operation of critical infrastructure, infringements of obligations under Union law intended to protect fundamental rights or serious damage to property or the environment.

+ (156) In order to ensure an appropriate and effective enforcement of the requirements and obligations set out by this Regulation, which is Union harmonisation legislation, the system of market surveillance and compliance of products established by Regulation (EU) 2019/1020 should apply in its entirety. Market surveillance authorities designated pursuant to this Regulation should have all enforcement powers laid down in this Regulation and in Regulation (EU) 2019/1020 and should exercise their powers and carry out their duties independently, impartially and without bias. Although the majority of AI systems are not subject to specific requirements and obligations under this Regulation, market surveillance authorities may take measures in relation to all AI systems when they present a risk in accordance with this Regulation. Due to the specific nature of Union institutions, agencies and bodies falling within the scope of this Regulation, it is appropriate to designate the European Data Protection Supervisor as a competent market surveillance authority for them. This should be without prejudice to the designation of national competent authorities by the Member States. Market surveillance activities should not affect the ability of the supervised entities to carry out their tasks independently, when such independence is required by Union law.

+ (157) This Regulation is without prejudice to the competences, tasks, powers and independence of relevant national public authorities or bodies which supervise the application of Union law protecting fundamental rights, including equality bodies and data protection authorities. Where necessary for their mandate, those national public authorities or bodies should also have access to any documentation created under this Regulation. A specific safeguard procedure should be set for ensuring adequate and timely enforcement against AI systems presenting a risk to health, safety and fundamental rights. The procedure for such AI systems presenting a risk should be applied to high-risk AI systems presenting a risk, prohibited systems which have been placed on the market, put into service or used in violation of the prohibited practices laid down in this Regulation and AI systems which have been made available in violation of the transparency requirements laid down in this Regulation and present a risk.

+ (158) Union financial services law includes internal governance and risk-management rules and requirements which are applicable to regulated financial institutions in the course of provision of those services, including when they make use of AI systems. In order to ensure coherent application and enforcement of the obligations under this Regulation and relevant rules and requirements of the Union financial services legal acts, the competent authorities for the supervision and enforcement of those legal acts, in particular competent authorities as defined in Regulation (EU) No 575/2013 of the European Parliament and of the Council (46) and Directives 2008/48/EC (47), 2009/138/EC (48),

+ (46) Regulation (EU) No 575/2013 of the European Parliament and of the Council of 26 June 2013 on prudential requirements for credit institutions and investment firms and amending Regulation (EU) No 648/2012 (OJ L 176, 27.6.2013, p. 1).

+ (47) Directive 2008/48/EC of the European Parliament and of the Council of 23 April 2008 on credit agreements for consumers and repealing Council Directive 87/102/EEC (OJ L 133, 22.5.2008, p. 66).

+ (48) Directive 2009/138/EC of the European Parliament and of the Council of 25 November 2009 on the taking-up and pursuit of the business of Insurance and Reinsurance (Solvency II) (OJ L 335, 17.12.2009, p. 1).

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 39/144

+ 2013/36/EU (49), 2014/17/EU (50) and (EU) 2016/97 (51) of the European Parliament and of the Council, should be designated, within their respective competences, as competent authorities for the purpose of supervising the implementation of this Regulation, including for market surveillance activities, as regards AI systems provided or used by regulated and supervised financial institutions unless Member States decide to designate another authority to fulfil these market surveillance tasks. Those competent authorities should have all powers under this Regulation and Regulation (EU) 2019/1020 to enforce the requirements and obligations of this Regulation, including powers to carry our ex post market surveillance activities that can be integrated, as appropriate, into their existing supervisory mechanisms and procedures under the relevant Union financial services law. It is appropriate to envisage that, when acting as market surveillance authorities under this Regulation, the national authorities responsible for the supervision of credit institutions regulated under Directive 2013/36/EU, which are participating in the Single Supervisory Mechanism established by Council Regulation (EU) No 1024/2013 (52), should report, without delay, to the European Central Bank any information identified in the course of their market surveillance activities that may be of potential interest for the European Central Bank's prudential supervisory tasks as specified in that Regulation. To further enhance the consistency between this Regulation and the rules applicable to credit institutions regulated under Directive 2013/36/EU, it is also appropriate to integrate some of the providers' procedural obligations in relation to risk management, post marketing monitoring and documentation into the existing obligations and procedures under Directive 2013/36/EU. In order to avoid overlaps, limited derogations should also be envisaged in relation to the quality management system of providers and the monitoring obligation placed on deployers of high-risk AI systems to the extent that these apply to credit institutions regulated by Directive 2013/36/EU. The same regime should apply to insurance and re-insurance undertakings and insurance holding companies under Directive 2009/138/EC and the insurance intermediaries under Directive (EU) 2016/97 and other types of financial institutions subject to requirements regarding internal governance, arrangements or processes established pursuant to the relevant Union financial services law to ensure consistency and equal treatment in the financial sector.

+ (159) Each market surveillance authority for high-risk AI systems in the area of biometrics, as listed in an annex to this Regulation insofar as those systems are used for the purposes of law enforcement, migration, asylum and border control management, or the administration of justice and democratic processes, should have effective investigative and corrective powers, including at least the power to obtain access to all personal data that are being processed and to all information necessary for the performance of its tasks. The market surveillance authorities should be able to exercise their powers by acting with complete independence. Any limitations of their access to sensitive operational data under this Regulation should be without prejudice to the powers conferred to them by Directive (EU) 2016/680. No exclusion on disclosing data to national data protection authorities under this Regulation should affect the current or future powers of those authorities beyond the scope of this Regulation.

+ (160) The market surveillance authorities and the Commission should be able to propose joint activities, including joint investigations, to be conducted by market surveillance authorities or market surveillance authorities jointly with the Commission, that have the aim of promoting compliance, identifying non-compliance, raising awareness and providing guidance in relation to this Regulation with respect to specific categories of high-risk AI systems that are found to present a serious risk across two or more Member States. Joint activities to promote compliance should be carried out in accordance with Article 9 of Regulation (EU) 2019/1020. The AI Office should provide coordination support for joint investigations.

+ (161) It is necessary to clarify the responsibilities and competences at Union and national level as regards AI systems that are built on general-purpose AI models. To avoid overlapping competences, where an AI system is based on a general-purpose AI model and the model and system are provided by the same provider, the supervision should

+ (49) Directive 2013/36/EU of the European Parliament and of the Council of 26 June 2013 on access to the activity of credit institutions and the prudential supervision of credit institutions and investment firms, amending Directive 2002/87/EC and repealing Directives 2006/48/EC and 2006/49/EC (OJ L 176, 27.6.2013, p. 338).

+ (50) Directive 2014/17/EU of the European Parliament and of the Council of 4 February 2014 on credit agreements for consumers relating to residential immovable property and amending Directives 2008/48/EC and 2013/36/EU and Regulation (EU) No 1093/2010 (OJ L 60, 28.2.2014, p. 34).

+ (51) Directive (EU) 2016/97 of the European Parliament and of the Council of 20 January 2016 on insurance distribution (OJ L 26, 2.2.2016, p. 19).

+ (52) Council Regulation (EU) No 1024/2013 of 15 October 2013 conferring specific tasks on the European Central Bank concerning policies relating to the prudential supervision of credit institutions (OJ L 287, 29.10.2013, p. 63).

+ 40/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ take place at Union level through the AI Office, which should have the powers of a market surveillance authority within the meaning of Regulation (EU) 2019/1020 for this purpose. In all other cases, national market surveillance authorities remain responsible for the supervision of AI systems. However, for general-purpose AI systems that can be used directly by deployers for at least one purpose that is classified as high-risk, market surveillance authorities should cooperate with the AI Office to carry out evaluations of compliance and inform the Board and other market surveillance authorities accordingly. Furthermore, market surveillance authorities should be able to request assistance from the AI Office where the market surveillance authority is unable to conclude an investigation on a high-risk AI system because of its inability to access certain information related to the general-purpose AI model on which the high-risk AI system is built. In such cases, the procedure regarding mutual assistance in cross-border cases in Chapter VI of Regulation (EU) 2019/1020 should apply mutatis mutandis.

+ (162) To make best use of the centralised Union expertise and synergies at Union level, the powers of supervision and enforcement of the obligations on providers of general-purpose AI models should be a competence of the Commission. The AI Office should be able to carry out all necessary actions to monitor the effective implementation of this Regulation as regards general-purpose AI models. It should be able to investigate possible infringements of the rules on providers of general-purpose AI models both on its own initiative, following the results of its monitoring activities, or upon request from market surveillance authorities in line with the conditions set out in this Regulation. To support effective monitoring of the AI Office, it should provide for the possibility that downstream providers lodge complaints about possible infringements of the rules on providers of general-purpose AI models and systems.

+ (163) With a view to complementing the governance systems for general-purpose AI models, the scientific panel should support the monitoring activities of the AI Office and may, in certain cases, provide qualified alerts to the AI Office which trigger follow-ups, such as investigations. This should be the case where the scientific panel has reason to suspect that a general-purpose AI model poses a concrete and identifiable risk at Union level. Furthermore, this should be the case where the scientific panel has reason to suspect that a general-purpose AI model meets the criteria that would lead to a classification as general-purpose AI model with systemic risk. To equip the scientific panel with the information necessary for the performance of those tasks, there should be a mechanism whereby the scientific panel can request the Commission to require documentation or information from a provider.

+ (164) The AI Office should be able to take the necessary actions to monitor the effective implementation of and compliance with the obligations for providers of general-purpose AI models laid down in this Regulation. The AI Office should be able to investigate possible infringements in accordance with the powers provided for in this Regulation, including by requesting documentation and information, by conducting evaluations, as well as by requesting measures from providers of general-purpose AI models. When conducting evaluations, in order to make use of independent expertise, the AI Office should be able to involve independent experts to carry out the evaluations on its behalf. Compliance with the obligations should be enforceable, inter alia, through requests to take appropriate measures, including risk mitigation measures in the case of identified systemic risks as well as restricting the making available on the market, withdrawing or recalling the model. As a safeguard, where needed beyond the procedural rights provided for in this Regulation, providers of general-purpose AI models should have the procedural rights provided for in Article 18 of Regulation (EU) 2019/1020, which should apply mutatis mutandis, without prejudice to more specific procedural rights provided for by this Regulation.

+ (165) The development of AI systems other than high-risk AI systems in accordance with the requirements of this Regulation may lead to a larger uptake of ethical and trustworthy AI in the Union. Providers of AI systems that are not high-risk should be encouraged to create codes of conduct, including related governance mechanisms, intended to foster the voluntary application of some or all of the mandatory requirements applicable to high-risk AI systems, adapted in light of the intended purpose of the systems and the lower risk involved and taking into account the available technical solutions and industry best practices such as model and data cards. Providers and, as appropriate, deployers of all AI systems, high-risk or not, and AI models should also be encouraged to apply on a voluntary basis additional requirements related, for example, to the elements of the Union's Ethics Guidelines for Trustworthy AI,

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 41/144

+ environmental sustainability, AI literacy measures, inclusive and diverse design and development of AI systems, including attention to vulnerable persons and accessibility to persons with disability, stakeholders' participation with the involvement, as appropriate, of relevant stakeholders such as business and civil society organisations, academia, research organisations, trade unions and consumer protection organisations in the design and development of AI systems, and diversity of the development teams, including gender balance. To ensure that the voluntary codes of conduct are effective, they should be based on clear objectives and key performance indicators to measure the achievement of those objectives. They should also be developed in an inclusive way, as appropriate, with the involvement of relevant stakeholders such as business and civil society organisations, academia, research organisations, trade unions and consumer protection organisation. The Commission may develop initiatives, including of a sectoral nature, to facilitate the lowering of technical barriers hindering cross-border exchange of data for AI development, including on data access infrastructure, semantic and technical interoperability of different types of data.

+ (166) It is important that AI systems related to products that are not high-risk in accordance with this Regulation and thus are not required to comply with the requirements set out for high-risk AI systems are nevertheless safe when placed on the market or put into service. To contribute to this objective, Regulation (EU) 2023/988 of the European Parliament and of the Council (53) would apply as a safety net.

+ (167) In order to ensure trustful and constructive cooperation of competent authorities on Union and national level, all parties involved in the application of this Regulation should respect the confidentiality of information and data obtained in carrying out their tasks, in accordance with Union or national law. They should carry out their tasks and activities in such a manner as to protect, in particular, intellectual property rights, confidential business information and trade secrets, the effective implementation of this Regulation, public and national security interests, the integrity of criminal and administrative proceedings, and the integrity of classified information.

+ (168) Compliance with this Regulation should be enforceable by means of the imposition of penalties and other enforcement measures. Member States should take all necessary measures to ensure that the provisions of this Regulation are implemented, including by laying down effective, proportionate and dissuasive penalties for their infringement, and to respect the ne bis in idem principle. In order to strengthen and harmonise administrative penalties for infringement of this Regulation, the upper limits for setting the administrative fines for certain specific infringements should be laid down. When assessing the amount of the fines, Member States should, in each individual case, take into account all relevant circumstances of the specific situation, with due regard in particular to the nature, gravity and duration of the infringement and of its consequences and to the size of the provider, in particular if the provider is an SME, including a start-up. The European Data Protection Supervisor should have the power to impose fines on Union institutions, agencies and bodies falling within the scope of this Regulation.

+ (169) Compliance with the obligations on providers of general-purpose AI models imposed under this Regulation should be enforceable, inter alia, by means of fines. To that end, appropriate levels of fines should also be laid down for infringement of those obligations, including the failure to comply with measures requested by the Commission in accordance with this Regulation, subject to appropriate limitation periods in accordance with the principle of proportionality. All decisions taken by the Commission under this Regulation are subject to review by the Court of Justice of the European Union in accordance with the TFEU, including the unlimited jurisdiction of the Court of Justice with regard to penalties pursuant to Article 261 TFEU.

+ (170) Union and national law already provide effective remedies to natural and legal persons whose rights and freedoms are adversely affected by the use of AI systems. Without prejudice to those remedies, any natural or legal person that has grounds to consider that there has been an infringement of this Regulation should be entitled to lodge a complaint to the relevant market surveillance authority.

+ (171) Affected persons should have the right to obtain an explanation where a deployer's decision is based mainly upon the output from certain high-risk AI systems that fall within the scope of this Regulation and where that decision produces legal effects or similarly significantly affects those persons in a way that they consider to have an adverse

+ (53) Regulation (EU) 2023/988 of the European Parliament and of the Council of 10 May 2023 on general product safety, amending Regulation (EU) No 1025/2012 of the European Parliament and of the Council and Directive (EU) 2020/1828 of the European Parliament and the Council, and repealing Directive 2001/95/EC of the European Parliament and of the Council and Council Directive 87/357/EEC (OJ L 135, 23.5.2023, p. 1).

+ 42/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ impact on their health, safety or fundamental rights. That explanation should be clear and meaningful and should provide a basis on which the affected persons are able to exercise their rights. The right to obtain an explanation should not apply to the use of AI systems for which exceptions or restrictions follow from Union or national law and should apply only to the extent this right is not already provided for under Union law.

+ (172) Persons acting as whistleblowers on the infringements of this Regulation should be protected under the Union law. Directive (EU) 2019/1937 of the European Parliament and of the Council (54) should therefore apply to the reporting of infringements of this Regulation and the protection of persons reporting such infringements.

+ (173) In order to ensure that the regulatory framework can be adapted where necessary, the power to adopt acts in accordance with Article 290 TFEU should be delegated to the Commission to amend the conditions under which an AI system is not to be considered to be high-risk, the list of high-risk AI systems, the provisions regarding technical documentation, the content of the EU declaration of conformity the provisions regarding the conformity assessment procedures, the provisions establishing the high-risk AI systems to which the conformity assessment procedure based on assessment of the quality management system and assessment of the technical documentation should apply, the threshold, benchmarks and indicators, including by supplementing those benchmarks and indicators, in the rules for the classification of general-purpose AI models with systemic risk, the criteria for the designation of general-purpose AI models with systemic risk, the technical documentation for providers of general-purpose AI models and the transparency information for providers of general-purpose AI models. It is of particular importance that the Commission carry out appropriate consultations during its preparatory work, including at expert level, and that those consultations be conducted in accordance with the principles laid down in the Interinstitutional Agreement of 13 April 2016 on Better Law-Making (55). In particular, to ensure equal participation in the preparation of delegated acts, the European Parliament and the Council receive all documents at the same time as Member States' experts, and their experts systematically have access to meetings of Commission expert groups dealing with the preparation of delegated acts.

+ (174) Given the rapid technological developments and the technical expertise required to effectively apply this Regulation, the Commission should evaluate and review this Regulation by 2 August 2029 and every four years thereafter and report to the European Parliament and the Council. In addition, taking into account the implications for the scope of this Regulation, the Commission should carry out an assessment of the need to amend the list of high-risk AI systems and the list of prohibited practices once a year. Moreover, by 2 August 2028 and every four years thereafter, the Commission should evaluate and report to the European Parliament and to the Council on the need to amend the list of high-risk areas headings in the annex to this Regulation, the AI systems within the scope of the transparency obligations, the effectiveness of the supervision and governance system and the progress on the development of standardisation deliverables on energy efficient development of general-purpose AI models, including the need for further measures or actions. Finally, by 2 August 2028 and every three years thereafter, the Commission should evaluate the impact and effectiveness of voluntary codes of conduct to foster the application of the requirements provided for high-risk AI systems in the case of AI systems other than high-risk AI systems and possibly other additional requirements for such AI systems.

+ (175) In order to ensure uniform conditions for the implementation of this Regulation, implementing powers should be conferred on the Commission. Those powers should be exercised in accordance with Regulation (EU) No 182/2011 of the European Parliament and of the Council (56).

+ (176) Since the objective of this Regulation, namely to improve the functioning of the internal market and to promote the uptake of human centric and trustworthy AI, while ensuring a high level of protection of health, safety, fundamental rights enshrined in the Charter, including democracy, the rule of law and environmental protection against harmful effects of AI systems in the Union and supporting innovation, cannot be sufficiently achieved by the Member States and can rather, by reason of the scale or effects of the action, be better achieved at Union level, the Union may adopt

+ (54) Directive (EU) 2019/1937 of the European Parliament and of the Council of 23 October 2019 on the protection of persons who report breaches of Union law (OJ L 305, 26.11.2019, p. 17).

+ (55) OJ L 123, 12.5.2016, p. 1. (56) Regulation (EU) No 182/2011 of the European Parliament and of the Council of 16 February 2011 laying down the rules and

+ general principles concerning mechanisms for control by Member States of the Commission's exercise of implementing powers (OJ L 55, 28.2.2011, p. 13).

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 43/144

+ measures in accordance with the principle of subsidiarity as set out in Article 5 TEU. In accordance with the principle of proportionality as set out in that Article, this Regulation does not go beyond what is necessary in order to achieve that objective.

+ (177) In order to ensure legal certainty, ensure an appropriate adaptation period for operators and avoid disruption to the market, including by ensuring continuity of the use of AI systems, it is appropriate that this Regulation applies to the high-risk AI systems that have been placed on the market or put into service before the general date of application thereof, only if, from that date, those systems are subject to significant changes in their design or intended purpose. It is appropriate to clarify that, in this respect, the concept of significant change should be understood as equivalent in substance to the notion of substantial modification, which is used with regard only to high-risk AI systems pursuant to this Regulation. On an exceptional basis and in light of public accountability, operators of AI systems which are components of the large-scale IT systems established by the legal acts listed in an annex to this Regulation and operators of high-risk AI systems that are intended to be used by public authorities should, respectively, take the necessary steps to comply with the requirements of this Regulation by end of 2030 and by 2 August 2030.

+ (178) Providers of high-risk AI systems are encouraged to start to comply, on a voluntary basis, with the relevant obligations of this Regulation already during the transitional period.

+ (179) This Regulation should apply from 2 August 2026. However, taking into account the unacceptable risk associated with the use of AI in certain ways, the prohibitions as well as the general provisions of this Regulation should already apply from 2 February 2025. While the full effect of those prohibitions follows with the establishment of the governance and enforcement of this Regulation, anticipating the application of the prohibitions is important to take account of unacceptable risks and to have an effect on other procedures, such as in civil law. Moreover, the infrastructure related to the governance and the conformity assessment system should be operational before 2 August 2026, therefore the provisions on notified bodies and governance structure should apply from 2 August 2025. Given the rapid pace of technological advancements and adoption of general-purpose AI models, obligations for providers of general-purpose AI models should apply from 2 August 2025. Codes of practice should be ready by 2 May 2025 in view of enabling providers to demonstrate compliance on time. The AI Office should ensure that classification rules and procedures are up to date in light of technological developments. In addition, Member States should lay down and notify to the Commission the rules on penalties, including administrative fines, and ensure that they are properly and effectively implemented by the date of application of this Regulation. Therefore the provisions on penalties should apply from 2 August 2025.

+ (180) The European Data Protection Supervisor and the European Data Protection Board were consulted in accordance with Article 42(1) and (2) of Regulation (EU) 2018/1725 and delivered their joint opinion on 18 June 2021,

+ HAVE ADOPTED THIS REGULATION:

+ CHAPTER I GENERAL PROVISIONS

+ Article 1 Subject matter`

+ 1. The purpose of this Regulation is to improve the functioning of the internal market and promote the uptake of human-centric and trustworthy artificial intelligence (AI), while ensuring a high level of protection of health, safety, fundamental rights enshrined in the Charter, including democracy, the rule of law and environmental protection, against the harmful effects of AI systems in the Union and supporting innovation.

+ 2. This Regulation lays down:

+ (a) harmonised rules for the placing on the market, the putting into service, and the use of AI systems in the Union;

+ 44/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ (b) prohibitions of certain AI practices;

+ (c) specific requirements for high-risk AI systems and obligations for operators of such systems;

+ (d) harmonised transparency rules for certain AI systems;

+ (e) harmonised rules for the placing on the market of general-purpose AI models;

+ (f) rules on market monitoring, market surveillance, governance and enforcement;

+ (g) measures to support innovation, with a particular focus on SMEs, including start-ups.

+ Article 2 Scope

+ 1. This Regulation applies to:

+ (a) providers placing on the market or putting into service AI systems or placing on the market general-purpose AI models in the Union, irrespective of whether those providers are established or located within the Union or in a third country;

+ (b) deployers of AI systems that have their place of establishment or are located within the Union;

+ (c) providers and deployers of AI systems that have their place of establishment or are located in a third country, where the output produced by the AI system is used in the Union;

+ (d) importers and distributors of AI systems;

+ (e) product manufacturers placing on the market or putting into service an AI system together with their product and under their own name or trademark;

+ (f) authorised representatives of providers, which are not established in the Union;

+ (g) affected persons that are located in the Union.

+ 2. For AI systems classified as high-risk AI systems in accordance with Article 6(1) related to products covered by the Union harmonisation legislation listed in Section B of Annex I, only Article 6(1), Articles 102 to 109 and Article 112 apply. Article 57 applies only in so far as the requirements for high-risk AI systems under this Regulation have been integrated in that Union harmonisation legislation.

+ 3. This Regulation does not apply to areas outside the scope of Union law, and shall not, in any event, affect the competences of the Member States concerning national security, regardless of the type of entity entrusted by the Member States with carrying out tasks in relation to those competences.

+ This Regulation does not apply to AI systems where and in so far they are placed on the market, put into service, or used with or without modification exclusively for military, defence or national security purposes, regardless of the type of entity carrying out those activities.

+ This Regulation does not apply to AI systems which are not placed on the market or put into service in the Union, where the output is used in the Union exclusively for military, defence or national security purposes, regardless of the type of entity carrying out those activities.

+ 4. This Regulation applies neither to public authorities in a third country nor to international organisations falling within the scope of this Regulation pursuant to paragraph 1, where those authorities or organisations use AI systems in the framework of international cooperation or agreements for law enforcement and judicial cooperation with the Union or with one or more Member States, provided that such a third country or international organisation provides adequate safeguards with respect to the protection of fundamental rights and freedoms of individuals.

+ 5. This Regulation shall not affect the application of the provisions on the liability of providers of intermediary services as set out in Chapter II of Regulation (EU) 2022/2065.

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 45/144

+ 6. This Regulation does not apply to AI systems or AI models, including their output, specifically developed and put into service for the sole purpose of scientific research and development.

+ 7. Union law on the protection of personal data, privacy and the confidentiality of communications applies to personal data processed in connection with the rights and obligations laid down in this Regulation. This Regulation shall not affect Regulation (EU) 2016/679 or (EU) 2018/1725, or Directive 2002/58/EC or (EU) 2016/680, without prejudice to Article 10(5) and Article 59 of this Regulation.

+ 8. This Regulation does not apply to any research, testing or development activity regarding AI systems or AI models prior to their being placed on the market or put into service. Such activities shall be conducted in accordance with applicable Union law. Testing in real world conditions shall not be covered by that exclusion.

+ 9. This Regulation is without prejudice to the rules laid down by other Union legal acts related to consumer protection and product safety.

+ 10. This Regulation does not apply to obligations of deployers who are natural persons using AI systems in the course of a purely personal non-professional activity.

+ 11. This Regulation does not preclude the Union or Member States from maintaining or introducing laws, regulations or administrative provisions which are more favourable to workers in terms of protecting their rights in respect of the use of AI systems by employers, or from encouraging or allowing the application of collective agreements which are more favourable to workers.

+ 12. This Regulation does not apply to AI systems released under free and open-source licences, unless they are placed on the market or put into service as high-risk AI systems or as an AI system that falls under Article 5 or 50.

+ Article 3 Definitions

+ For the purposes of this Regulation, the following definitions apply:

+ (1) `AI system' means a machine-based system that is designed to operate with varying levels of autonomy and that may exhibit adaptiveness after deployment, and that, for explicit or implicit objectives, infers, from the input it receives, how to generate outputs such as predictions, content, recommendations, or decisions that can influence physical or virtual environments;

+ (2) `risk' means the combination of the probability of an occurrence of harm and the severity of that harm;

+ (3) `provider' means a natural or legal person, public authority, agency or other body that develops an AI system or a general-purpose AI model or that has an AI system or a general-purpose AI model developed and places it on the market or puts the AI system into service under its own name or trademark, whether for payment or free of charge;

+ (4) `deployer' means a natural or legal person, public authority, agency or other body using an AI system under its authority except where the AI system is used in the course of a personal non-professional activity;

+ (5) `authorised representative' means a natural or legal person located or established in the Union who has received and accepted a written mandate from a provider of an AI system or a general-purpose AI model to, respectively, perform and carry out on its behalf the obligations and procedures established by this Regulation;

+ (6) `importer' means a natural or legal person located or established in the Union that places on the market an AI system that bears the name or trademark of a natural or legal person established in a third country;

+ (7) `distributor' means a natural or legal person in the supply chain, other than the provider or the importer, that makes an AI system available on the Union market;

+ (8) `operator' means a provider, product manufacturer, deployer, authorised representative, importer or distributor;

+ 46/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ (9) `placing on the market' means the first making available of an AI system or a general-purpose AI model on the Union market;

+ (10) `making available on the market' means the supply of an AI system or a general-purpose AI model for distribution or use on the Union market in the course of a commercial activity, whether in return for payment or free of charge;

+ (11) `putting into service' means the supply of an AI system for first use directly to the deployer or for own use in the Union for its intended purpose;

+ (12) `intended purpose' means the use for which an AI system is intended by the provider, including the specific context and conditions of use, as specified in the information supplied by the provider in the instructions for use, promotional or sales materials and statements, as well as in the technical documentation;

+ (13) `reasonably foreseeable misuse' means the use of an AI system in a way that is not in accordance with its intended purpose, but which may result from reasonably foreseeable human behaviour or interaction with other systems, including other AI systems;

+ (14) `safety component' means a component of a product or of an AI system which fulfils a safety function for that product or AI system, or the failure or malfunctioning of which endangers the health and safety of persons or property;

+ (15) `instructions for use' means the information provided by the provider to inform the deployer of, in particular, an AI system's intended purpose and proper use;

+ (16) `recall of an AI system' means any measure aiming to achieve the return to the provider or taking out of service or disabling the use of an AI system made available to deployers;

+ (17) `withdrawal of an AI system' means any measure aiming to prevent an AI system in the supply chain being made available on the market;

+ (18) `performance of an AI system' means the ability of an AI system to achieve its intended purpose;

+ (19) `notifying authority' means the national authority responsible for setting up and carrying out the necessary procedures for the assessment, designation and notification of conformity assessment bodies and for their monitoring;

+ (20) `conformity assessment' means the process of demonstrating whether the requirements set out in Chapter III, Section 2 relating to a high-risk AI system have been fulfilled;

+ (21) `conformity assessment body' means a body that performs third-party conformity assessment activities, including testing, certification and inspection;

+ (22) `notified body' means a conformity assessment body notified in accordance with this Regulation and other relevant Union harmonisation legislation;

+ (23) `substantial modification' means a change to an AI system after its placing on the market or putting into service which is not foreseen or planned in the initial conformity assessment carried out by the provider and as a result of which the compliance of the AI system with the requirements set out in Chapter III, Section 2 is affected or results in a modification to the intended purpose for which the AI system has been assessed;

+ (24) `CE marking' means a marking by which a provider indicates that an AI system is in conformity with the requirements set out in Chapter III, Section 2 and other applicable Union harmonisation legislation providing for its affixing;

+ (25) `post-market monitoring system' means all activities carried out by providers of AI systems to collect and review experience gained from the use of AI systems they place on the market or put into service for the purpose of identifying any need to immediately apply any necessary corrective or preventive actions;

+ (26) `market surveillance authority' means the national authority carrying out the activities and taking the measures pursuant to Regulation (EU) 2019/1020;

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 47/144

+ (27) `harmonised standard' means a harmonised standard as defined in Article 2(1), point (c), of Regulation (EU) No 1025/2012;

+ (28) `common specification' means a set of technical specifications as defined in Article 2, point (4) of Regulation (EU) No 1025/2012, providing means to comply with certain requirements established under this Regulation;

+ (29) `training data' means data used for training an AI system through fitting its learnable parameters;

+ (30) `validation data' means data used for providing an evaluation of the trained AI system and for tuning its non-learnable parameters and its learning process in order, inter alia, to prevent underfitting or overfitting;

+ (31) `validation data set' means a separate data set or part of the training data set, either as a fixed or variable split;

+ (32) `testing data' means data used for providing an independent evaluation of the AI system in order to confirm the expected performance of that system before its placing on the market or putting into service;

+ (33) `input data' means data provided to or directly acquired by an AI system on the basis of which the system produces an output;

+ (34) `biometric data' means personal data resulting from specific technical processing relating to the physical, physiological or behavioural characteristics of a natural person, such as facial images or dactyloscopic data;

+ (35) `biometric identification' means the automated recognition of physical, physiological, behavioural, or psychological human features for the purpose of establishing the identity of a natural person by comparing biometric data of that individual to biometric data of individuals stored in a database;

+ (36) `biometric verification' means the automated, one-to-one verification, including authentication, of the identity of natural persons by comparing their biometric data to previously provided biometric data;

+ (37) `special categories of personal data' means the categories of personal data referred to in Article 9(1) of Regulation (EU) 2016/679, Article 10 of Directive (EU) 2016/680 and Article 10(1) of Regulation (EU) 2018/1725;

+ (38) `sensitive operational data' means operational data related to activities of prevention, detection, investigation or prosecution of criminal offences, the disclosure of which could jeopardise the integrity of criminal proceedings;

+ (39) `emotion recognition system' means an AI system for the purpose of identifying or inferring emotions or intentions of natural persons on the basis of their biometric data;

+ (40) `biometric categorisation system' means an AI system for the purpose of assigning natural persons to specific categories on the basis of their biometric data, unless it is ancillary to another commercial service and strictly necessary for objective technical reasons;

+ (41) `remote biometric identification system' means an AI system for the purpose of identifying natural persons, without their active involvement, typically at a distance through the comparison of a person's biometric data with the biometric data contained in a reference database;

+ (42) `real-time remote biometric identification system' means a remote biometric identification system, whereby the capturing of biometric data, the comparison and the identification all occur without a significant delay, comprising not only instant identification, but also limited short delays in order to avoid circumvention;

+ (43) `post-remote biometric identification system' means a remote biometric identification system other than a real-time remote biometric identification system;

+ (44) `publicly accessible space' means any publicly or privately owned physical place accessible to an undetermined number of natural persons, regardless of whether certain conditions for access may apply, and regardless of the potential capacity restrictions;

+ 48/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ (45) `law enforcement authority' means:

+ (a) any public authority competent for the prevention, investigation, detection or prosecution of criminal offences or the execution of criminal penalties, including the safeguarding against and the prevention of threats to public security; or

+ (b) any other body or entity entrusted by Member State law to exercise public authority and public powers for the purposes of the prevention, investigation, detection or prosecution of criminal offences or the execution of criminal penalties, including the safeguarding against and the prevention of threats to public security;

+ (46) `law enforcement' means activities carried out by law enforcement authorities or on their behalf for the prevention, investigation, detection or prosecution of criminal offences or the execution of criminal penalties, including safeguarding against and preventing threats to public security;

+ (47) `AI Office' means the Commission's function of contributing to the implementation, monitoring and supervision of AI systems and general-purpose AI models, and AI governance, provided for in Commission Decision of 24 January 2024; references in this Regulation to the AI Office shall be construed as references to the Commission;

+ (48) `national competent authority' means a notifying authority or a market surveillance authority; as regards AI systems put into service or used by Union institutions, agencies, offices and bodies, references to national competent authorities or market surveillance authorities in this Regulation shall be construed as references to the European Data Protection Supervisor;

+ (49) `serious incident' means an incident or malfunctioning of an AI system that directly or indirectly leads to any of the following:

+ (a) the death of a person, or serious harm to a person's health;

+ (b) a serious and irreversible disruption of the management or operation of critical infrastructure;

+ (c) the infringement of obligations under Union law intended to protect fundamental rights;

+ (d) serious harm to property or the environment;

+ (50) `personal data' means personal data as defined in Article 4, point (1), of Regulation (EU) 2016/679;

+ (51) `non-personal data' means data other than personal data as defined in Article 4, point (1), of Regulation (EU) 2016/679;

+ (52) `profiling' means profiling as defined in Article 4, point (4), of Regulation (EU) 2016/679;

+ (53) `real-world testing plan' means a document that describes the objectives, methodology, geographical, population and temporal scope, monitoring, organisation and conduct of testing in real-world conditions;

+ (54) `sandbox plan' means a document agreed between the participating provider and the competent authority describing the objectives, conditions, timeframe, methodology and requirements for the activities carried out within the sandbox;

+ (55) `AI regulatory sandbox' means a controlled framework set up by a competent authority which offers providers or prospective providers of AI systems the possibility to develop, train, validate and test, where appropriate in real-world conditions, an innovative AI system, pursuant to a sandbox plan for a limited time under regulatory supervision;

+ (56) `AI literacy' means skills, knowledge and understanding that allow providers, deployers and affected persons, taking into account their respective rights and obligations in the context of this Regulation, to make an informed deployment of AI systems, as well as to gain awareness about the opportunities and risks of AI and possible harm it can cause;

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 49/144

+ (57) `testing in real-world conditions' means the temporary testing of an AI system for its intended purpose in real-world conditions outside a laboratory or otherwise simulated environment, with a view to gathering reliable and robust data and to assessing and verifying the conformity of the AI system with the requirements of this Regulation and it does not qualify as placing the AI system on the market or putting it into service within the meaning of this Regulation, provided that all the conditions laid down in Article 57 or 60 are fulfilled;

+ (58) `subject', for the purpose of real-world testing, means a natural person who participates in testing in real-world conditions;

+ (59) `informed consent' means a subject's freely given, specific, unambiguous and voluntary expression of his or her willingness to participate in a particular testing in real-world conditions, after having been informed of all aspects of the testing that are relevant to the subject's decision to participate;

+ (60) `deep fake' means AI-generated or manipulated image, audio or video content that resembles existing persons, objects, places, entities or events and would falsely appear to a person to be authentic or truthful;

+ (61) `widespread infringement' means any act or omission contrary to Union law protecting the interest of individuals, which:

+ (a) has harmed or is likely to harm the collective interests of individuals residing in at least two Member States other than the Member State in which:

+ (i) the act or omission originated or took place;

+ (ii) the provider concerned, or, where applicable, its authorised representative is located or established; or

+ (iii) the deployer is established, when the infringement is committed by the deployer;

+ (b) has caused, causes or is likely to cause harm to the collective interests of individuals and has common features, including the same unlawful practice or the same interest being infringed, and is occurring concurrently, committed by the same operator, in at least three Member States;

+ (62) `critical infrastructure' means critical infrastructure as defined in Article 2, point (4), of Directive (EU) 2022/2557;

+ (63) `general-purpose AI model' means an AI model, including where such an AI model is trained with a large amount of data using self-supervision at scale, that displays significant generality and is capable of competently performing a wide range of distinct tasks regardless of the way the model is placed on the market and that can be integrated into a variety of downstream systems or applications, except AI models that are used for research, development or prototyping activities before they are placed on the market;

+ (64) `high-impact capabilities' means capabilities that match or exceed the capabilities recorded in the most advanced general-purpose AI models;

+ (65) `systemic risk' means a risk that is specific to the high-impact capabilities of general-purpose AI models, having a significant impact on the Union market due to their reach, or due to actual or reasonably foreseeable negative effects on public health, safety, public security, fundamental rights, or the society as a whole, that can be propagated at scale across the value chain;

+ (66) `general-purpose AI system' means an AI system which is based on a general-purpose AI model and which has the capability to serve a variety of purposes, both for direct use as well as for integration in other AI systems;

+ (67) `floating-point operation' means any mathematical operation or assignment involving floating-point numbers, which are a subset of the real numbers typically represented on computers by an integer of fixed precision scaled by an integer exponent of a fixed base;

+ (68) `downstream provider' means a provider of an AI system, including a general-purpose AI system, which integrates an AI model, regardless of whether the AI model is provided by themselves and vertically integrated or provided by another entity based on contractual relations.

+ 50/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ Article 4 AI literacy

+ Providers and deployers of AI systems shall take measures to ensure, to their best extent, a sufficient level of AI literacy of their staff and other persons dealing with the operation and use of AI systems on their behalf, taking into account their technical knowledge, experience, education and training and the context the AI systems are to be used in, and considering the persons or groups of persons on whom the AI systems are to be used.

+ CHAPTER II PROHIBITED AI PRACTICES

+ Article 5 Prohibited AI practices

+ 1. The following AI practices shall be prohibited:

+ (a) the placing on the market, the putting into service or the use of an AI system that deploys subliminal techniques beyond a person's consciousness or purposefully manipulative or deceptive techniques, with the objective, or the effect of materially distorting the behaviour of a person or a group of persons by appreciably impairing their ability to make an informed decision, thereby causing them to take a decision that they would not have otherwise taken in a manner that causes or is reasonably likely to cause that person, another person or group of persons significant harm;

+ (b) the placing on the market, the putting into service or the use of an AI system that exploits any of the vulnerabilities of a natural person or a specific group of persons due to their age, disability or a specific social or economic situation, with the objective, or the effect, of materially distorting the behaviour of that person or a person belonging to that group in a manner that causes or is reasonably likely to cause that person or another person significant harm;

+ (c) the placing on the market, the putting into service or the use of AI systems for the evaluation or classification of natural persons or groups of persons over a certain period of time based on their social behaviour or known, inferred or predicted personal or personality characteristics, with the social score leading to either or both of the following:

+ (i) detrimental or unfavourable treatment of certain natural persons or groups of persons in social contexts that are unrelated to the contexts in which the data was originally generated or collected;

+ (ii) detrimental or unfavourable treatment of certain natural persons or groups of persons that is unjustified or disproportionate to their social behaviour or its gravity;

+ (d) the placing on the market, the putting into service for this specific purpose, or the use of an AI system for making risk assessments of natural persons in order to assess or predict the risk of a natural person committing a criminal offence, based solely on the profiling of a natural person or on assessing their personality traits and characteristics; this prohibition shall not apply to AI systems used to support the human assessment of the involvement of a person in a criminal activity, which is already based on objective and verifiable facts directly linked to a criminal activity;

+ (e) the placing on the market, the putting into service for this specific purpose, or the use of AI systems that create or expand facial recognition databases through the untargeted scraping of facial images from the internet or CCTV footage;

+ (f) the placing on the market, the putting into service for this specific purpose, or the use of AI systems to infer emotions of a natural person in the areas of workplace and education institutions, except where the use of the AI system is intended to be put in place or into the market for medical or safety reasons;

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 51/144

+ (g) the placing on the market, the putting into service for this specific purpose, or the use of biometric categorisation systems that categorise individually natural persons based on their biometric data to deduce or infer their race, political opinions, trade union membership, religious or philosophical beliefs, sex life or sexual orientation; this prohibition does not cover any labelling or filtering of lawfully acquired biometric datasets, such as images, based on biometric data or categorizing of biometric data in the area of law enforcement;

+ (h) the use of `real-time' remote biometric identification systems in publicly accessible spaces for the purposes of law enforcement, unless and in so far as such use is strictly necessary for one of the following objectives:

+ (i) the targeted search for specific victims of abduction, trafficking in human beings or sexual exploitation of human beings, as well as the search for missing persons;

+ (ii) the prevention of a specific, substantial and imminent threat to the life or physical safety of natural persons or a genuine and present or genuine and foreseeable threat of a terrorist attack;

+ (iii) the localisation or identification of a person suspected of having committed a criminal offence, for the purpose of conducting a criminal investigation or prosecution or executing a criminal penalty for offences referred to in Annex II and punishable in the Member State concerned by a custodial sentence or a detention order for a maximum period of at least four years.

+ Point (h) of the first subparagraph is without prejudice to Article 9 of Regulation (EU) 2016/679 for the processing of biometric data for purposes other than law enforcement.

+ 2. The use of `real-time' remote biometric identification systems in publicly accessible spaces for the purposes of law enforcement for any of the objectives referred to in paragraph 1, first subparagraph, point (h), shall be deployed for the purposes set out in that point only to confirm the identity of the specifically targeted individual, and it shall take into account the following elements:

+ (a) the nature of the situation giving rise to the possible use, in particular the seriousness, probability and scale of the harm that would be caused if the system were not used;

+ (b) the consequences of the use of the system for the rights and freedoms of all persons concerned, in particular the seriousness, probability and scale of those consequences.

+ In addition, the use of `real-time' remote biometric identification systems in publicly accessible spaces for the purposes of law enforcement for any of the objectives referred to in paragraph 1, first subparagraph, point (h), of this Article shall comply with necessary and proportionate safeguards and conditions in relation to the use in accordance with the national law authorising the use thereof, in particular as regards the temporal, geographic and personal limitations. The use of the `real-time' remote biometric identification system in publicly accessible spaces shall be authorised only if the law enforcement authority has completed a fundamental rights impact assessment as provided for in Article 27 and has registered the system in the EU database according to Article 49. However, in duly justified cases of urgency, the use of such systems may be commenced without the registration in the EU database, provided that such registration is completed without undue delay.

+ 3. For the purposes of paragraph 1, first subparagraph, point (h) and paragraph 2, each use for the purposes of law enforcement of a `real-time' remote biometric identification system in publicly accessible spaces shall be subject to a prior authorisation granted by a judicial authority or an independent administrative authority whose decision is binding of the Member State in which the use is to take place, issued upon a reasoned request and in accordance with the detailed rules of national law referred to in paragraph 5. However, in a duly justified situation of urgency, the use of such system may be commenced without an authorisation provided that such authorisation is requested without undue delay, at the latest within 24 hours. If such authorisation is rejected, the use shall be stopped with immediate effect and all the data, as well as the results and outputs of that use shall be immediately discarded and deleted.

+ The competent judicial authority or an independent administrative authority whose decision is binding shall grant the authorisation only where it is satisfied, on the basis of objective evidence or clear indications presented to it, that the use of the `real-time' remote biometric identification system concerned is necessary for, and proportionate to, achieving one of the

+ 52/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ objectives specified in paragraph 1, first subparagraph, point (h), as identified in the request and, in particular, remains limited to what is strictly necessary concerning the period of time as well as the geographic and personal scope. In deciding on the request, that authority shall take into account the elements referred to in paragraph 2. No decision that produces an adverse legal effect on a person may be taken based solely on the output of the `real-time' remote biometric identification system.

+ 4. Without prejudice to paragraph 3, each use of a `real-time' remote biometric identification system in publicly accessible spaces for law enforcement purposes shall be notified to the relevant market surveillance authority and the national data protection authority in accordance with the national rules referred to in paragraph 5. The notification shall, as a minimum, contain the information specified under paragraph 6 and shall not include sensitive operational data.

+ 5. A Member State may decide to provide for the possibility to fully or partially authorise the use of `real-time' remote biometric identification systems in publicly accessible spaces for the purposes of law enforcement within the limits and under the conditions listed in paragraph 1, first subparagraph, point (h), and paragraphs 2 and 3. Member States concerned shall lay down in their national law the necessary detailed rules for the request, issuance and exercise of, as well as supervision and reporting relating to, the authorisations referred to in paragraph 3. Those rules shall also specify in respect of which of the objectives listed in paragraph 1, first subparagraph, point (h), including which of the criminal offences referred to in point (h)(iii) thereof, the competent authorities may be authorised to use those systems for the purposes of law enforcement. Member States shall notify those rules to the Commission at the latest 30 days following the adoption thereof. Member States may introduce, in accordance with Union law, more restrictive laws on the use of remote biometric identification systems.

+ 6. National market surveillance authorities and the national data protection authorities of Member States that have been notified of the use of `real-time' remote biometric identification systems in publicly accessible spaces for law enforcement purposes pursuant to paragraph 4 shall submit to the Commission annual reports on such use. For that purpose, the Commission shall provide Member States and national market surveillance and data protection authorities with a template, including information on the number of the decisions taken by competent judicial authorities or an independent administrative authority whose decision is binding upon requests for authorisations in accordance with paragraph 3 and their result.

+ 7. The Commission shall publish annual reports on the use of real-time remote biometric identification systems in publicly accessible spaces for law enforcement purposes, based on aggregated data in Member States on the basis of the annual reports referred to in paragraph 6. Those annual reports shall not include sensitive operational data of the related law enforcement activities.

+ 8. This Article shall not affect the prohibitions that apply where an AI practice infringes other Union law.

+ CHAPTER III HIGH-RISK AI SYSTEMS

+ SECTION 1

+ Classification of AI systems as high-risk

+ Article 6 Classification rules for high-risk AI systems

+ 1. Irrespective of whether an AI system is placed on the market or put into service independently of the products referred to in points (a) and (b), that AI system shall be considered to be high-risk where both of the following conditions are fulfilled:

+ (a) the AI system is intended to be used as a safety component of a product, or the AI system is itself a product, covered by the Union harmonisation legislation listed in Annex I;

+ (b) the product whose safety component pursuant to point (a) is the AI system, or the AI system itself as a product, is required to undergo a third-party conformity assessment, with a view to the placing on the market or the putting into service of that product pursuant to the Union harmonisation legislation listed in Annex I.

+ 2. In addition to the high-risk AI systems referred to in paragraph 1, AI systems referred to in Annex III shall be considered to be high-risk.

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 53/144

+ 3. By derogation from paragraph 2, an AI system referred to in Annex III shall not be considered to be high-risk where it does not pose a significant risk of harm to the health, safety or fundamental rights of natural persons, including by not materially influencing the outcome of decision making.

+ The first subparagraph shall apply where any of the following conditions is fulfilled:

+ (a) the AI system is intended to perform a narrow procedural task;

+ (b) the AI system is intended to improve the result of a previously completed human activity;

+ (c) the AI system is intended to detect decision-making patterns or deviations from prior decision-making patterns and is not meant to replace or influence the previously completed human assessment, without proper human review; or

+ (d) the AI system is intended to perform a preparatory task to an assessment relevant for the purposes of the use cases listed in Annex III.

+ Notwithstanding the first subparagraph, an AI system referred to in Annex III shall always be considered to be high-risk where the AI system performs profiling of natural persons.

+ 4. A provider who considers that an AI system referred to in Annex III is not high-risk shall document its assessment before that system is placed on the market or put into service. Such provider shall be subject to the registration obligation set out in Article 49(2). Upon request of national competent authorities, the provider shall provide the documentation of the assessment.

+ 5. The Commission shall, after consulting the European Artificial Intelligence Board (the `Board'), and no later than 2 February 2026, provide guidelines specifying the practical implementation of this Article in line with Article 96 together with a comprehensive list of practical examples of use cases of AI systems that are high-risk and not high-risk.

+ 6. The Commission is empowered to adopt delegated acts in accordance with Article 97 in order to amend paragraph 3, second subparagraph, of this Article by adding new conditions to those laid down therein, or by modifying them, where there is concrete and reliable evidence of the existence of AI systems that fall under the scope of Annex III, but do not pose a significant risk of harm to the health, safety or fundamental rights of natural persons.

+ 7. The Commission shall adopt delegated acts in accordance with Article 97 in order to amend paragraph 3, second subparagraph, of this Article by deleting any of the conditions laid down therein, where there is concrete and reliable evidence that this is necessary to maintain the level of protection of health, safety and fundamental rights provided for by this Regulation.

+ 8. Any amendment to the conditions laid down in paragraph 3, second subparagraph, adopted in accordance with paragraphs 6 and 7 of this Article shall not decrease the overall level of protection of health, safety and fundamental rights provided for by this Regulation and shall ensure consistency with the delegated acts adopted pursuant to Article 7(1), and take account of market and technological developments.

+ Article 7 Amendments to Annex III

+ 1. The Commission is empowered to adopt delegated acts in accordance with Article 97 to amend Annex III by adding or modifying use-cases of high-risk AI systems where both of the following conditions are fulfilled:

+ (a) the AI systems are intended to be used in any of the areas listed in Annex III;

+ (b) the AI systems pose a risk of harm to health and safety, or an adverse impact on fundamental rights, and that risk is equivalent to, or greater than, the risk of harm or of adverse impact posed by the high-risk AI systems already referred to in Annex III.

+ 54/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 2. When assessing the condition under paragraph 1, point (b), the Commission shall take into account the following criteria:

+ (a) the intended purpose of the AI system;

+ (b) the extent to which an AI system has been used or is likely to be used;

+ (c) the nature and amount of the data processed and used by the AI system, in particular whether special categories of personal data are processed;

+ (d) the extent to which the AI system acts autonomously and the possibility for a human to override a decision or recommendations that may lead to potential harm;

+ (e) the extent to which the use of an AI system has already caused harm to health and safety, has had an adverse impact on fundamental rights or has given rise to significant concerns in relation to the likelihood of such harm or adverse impact, as demonstrated, for example, by reports or documented allegations submitted to national competent authorities or by other reports, as appropriate;

+ (f) the potential extent of such harm or such adverse impact, in particular in terms of its intensity and its ability to affect multiple persons or to disproportionately affect a particular group of persons;

+ (g) the extent to which persons who are potentially harmed or suffer an adverse impact are dependent on the outcome produced with an AI system, in particular because for practical or legal reasons it is not reasonably possible to opt-out from that outcome;

+ (h) the extent to which there is an imbalance of power, or the persons who are potentially harmed or suffer an adverse impact are in a vulnerable position in relation to the deployer of an AI system, in particular due to status, authority, knowledge, economic or social circumstances, or age;

+ (i) the extent to which the outcome produced involving an AI system is easily corrigible or reversible, taking into account the technical solutions available to correct or reverse it, whereby outcomes having an adverse impact on health, safety or fundamental rights, shall not be considered to be easily corrigible or reversible;

+ (j) the magnitude and likelihood of benefit of the deployment of the AI system for individuals, groups, or society at large, including possible improvements in product safety;

+ (k) the extent to which existing Union law provides for:

+ (i) effective measures of redress in relation to the risks posed by an AI system, with the exclusion of claims for damages;

+ (ii) effective measures to prevent or substantially minimise those risks.

+ 3. The Commission is empowered to adopt delegated acts in accordance with Article 97 to amend the list in Annex III by removing high-risk AI systems where both of the following conditions are fulfilled:

+ (a) the high-risk AI system concerned no longer poses any significant risks to fundamental rights, health or safety, taking into account the criteria listed in paragraph 2;

+ (b) the deletion does not decrease the overall level of protection of health, safety and fundamental rights under Union law.

+ SECTION 2

+ Requirements for high-risk AI systems

+ Article 8 Compliance with the requirements

+ 1. High-risk AI systems shall comply with the requirements laid down in this Section, taking into account their intended purpose as well as the generally acknowledged state of the art on AI and AI-related technologies. The risk management system referred to in Article 9 shall be taken into account when ensuring compliance with those requirements.

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 55/144

+ 2. Where a product contains an AI system, to which the requirements of this Regulation as well as requirements of the Union harmonisation legislation listed in Section A of Annex I apply, providers shall be responsible for ensuring that their product is fully compliant with all applicable requirements under applicable Union harmonisation legislation. In ensuring the compliance of high-risk AI systems referred to in paragraph 1 with the requirements set out in this Section, and in order to ensure consistency, avoid duplication and minimise additional burdens, providers shall have a choice of integrating, as appropriate, the necessary testing and reporting processes, information and documentation they provide with regard to their product into documentation and procedures that already exist and are required under the Union harmonisation legislation listed in Section A of Annex I.

+ Article 9 Risk management system

+ 1. A risk management system shall be established, implemented, documented and maintained in relation to high-risk AI systems.

+ 2. The risk management system shall be understood as a continuous iterative process planned and run throughout the entire lifecycle of a high-risk AI system, requiring regular systematic review and updating. It shall comprise the following steps:

+ (a) the identification and analysis of the known and the reasonably foreseeable risks that the high-risk AI system can pose to health, safety or fundamental rights when the high-risk AI system is used in accordance with its intended purpose;

+ (b) the estimation and evaluation of the risks that may emerge when the high-risk AI system is used in accordance with its intended purpose, and under conditions of reasonably foreseeable misuse;

+ (c) the evaluation of other risks possibly arising, based on the analysis of data gathered from the post-market monitoring system referred to in Article 72;

+ (d) the adoption of appropriate and targeted risk management measures designed to address the risks identified pursuant to point (a).

+ 3. The risks referred to in this Article shall concern only those which may be reasonably mitigated or eliminated through the development or design of the high-risk AI system, or the provision of adequate technical information.

+ 4. The risk management measures referred to in paragraph 2, point (d), shall give due consideration to the effects and possible interaction resulting from the combined application of the requirements set out in this Section, with a view to minimising risks more effectively while achieving an appropriate balance in implementing the measures to fulfil those requirements.

+ 5. The risk management measures referred to in paragraph 2, point (d), shall be such that the relevant residual risk associated with each hazard, as well as the overall residual risk of the high-risk AI systems is judged to be acceptable.

+ In identifying the most appropriate risk management measures, the following shall be ensured:

+ (a) elimination or reduction of risks identified and evaluated pursuant to paragraph 2 in as far as technically feasible through adequate design and development of the high-risk AI system;

+ (b) where appropriate, implementation of adequate mitigation and control measures addressing risks that cannot be eliminated;

+ (c) provision of information required pursuant to Article 13 and, where appropriate, training to deployers.

+ With a view to eliminating or reducing risks related to the use of the high-risk AI system, due consideration shall be given to the technical knowledge, experience, education, the training to be expected by the deployer, and the presumable context in which the system is intended to be used.

+ 56/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 6. High-risk AI systems shall be tested for the purpose of identifying the most appropriate and targeted risk management measures. Testing shall ensure that high-risk AI systems perform consistently for their intended purpose and that they are in compliance with the requirements set out in this Section.

+ 7. Testing procedures may include testing in real-world conditions in accordance with Article 60.

+ 8. The testing of high-risk AI systems shall be performed, as appropriate, at any time throughout the development process, and, in any event, prior to their being placed on the market or put into service. Testing shall be carried out against prior defined metrics and probabilistic thresholds that are appropriate to the intended purpose of the high-risk AI system.

+ 9. When implementing the risk management system as provided for in paragraphs 1 to 7, providers shall give consideration to whether in view of its intended purpose the high-risk AI system is likely to have an adverse impact on persons under the age of 18 and, as appropriate, other vulnerable groups.

+ 10. For providers of high-risk AI systems that are subject to requirements regarding internal risk management processes under other relevant provisions of Union law, the aspects provided in paragraphs 1 to 9 may be part of, or combined with, the risk management procedures established pursuant to that law.

+ Article 10 Data and data governance

+ 1. High-risk AI systems which make use of techniques involving the training of AI models with data shall be developed on the basis of training, validation and testing data sets that meet the quality criteria referred to in paragraphs 2 to 5 whenever such data sets are used.

+ 2. Training, validation and testing data sets shall be subject to data governance and management practices appropriate for the intended purpose of the high-risk AI system. Those practices shall concern in particular:

+ (a) the relevant design choices;

+ (b) data collection processes and the origin of data, and in the case of personal data, the original purpose of the data collection;

+ (c) relevant data-preparation processing operations, such as annotation, labelling, cleaning, updating, enrichment and aggregation;

+ (d) the formulation of assumptions, in particular with respect to the information that the data are supposed to measure and represent;

+ (e) an assessment of the availability, quantity and suitability of the data sets that are needed;

+ (f) examination in view of possible biases that are likely to affect the health and safety of persons, have a negative impact on fundamental rights or lead to discrimination prohibited under Union law, especially where data outputs influence inputs for future operations;

+ (g) appropriate measures to detect, prevent and mitigate possible biases identified according to point (f);

+ (h) the identification of relevant data gaps or shortcomings that prevent compliance with this Regulation, and how those gaps and shortcomings can be addressed.

+ 3. Training, validation and testing data sets shall be relevant, sufficiently representative, and to the best extent possible, free of errors and complete in view of the intended purpose. They shall have the appropriate statistical properties, including, where applicable, as regards the persons or groups of persons in relation to whom the high-risk AI system is intended to be used. Those characteristics of the data sets may be met at the level of individual data sets or at the level of a combination thereof.

+ 4. Data sets shall take into account, to the extent required by the intended purpose, the characteristics or elements that are particular to the specific geographical, contextual, behavioural or functional setting within which the high-risk AI system is intended to be used.

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 57/144

+ 5. To the extent that it is strictly necessary for the purpose of ensuring bias detection and correction in relation to the high-risk AI systems in accordance with paragraph (2), points (f) and (g) of this Article, the providers of such systems may exceptionally process special categories of personal data, subject to appropriate safeguards for the fundamental rights and freedoms of natural persons. In addition to the provisions set out in Regulations (EU) 2016/679 and (EU) 2018/1725 and Directive (EU) 2016/680, all the following conditions must be met in order for such processing to occur:

+ (a) the bias detection and correction cannot be effectively fulfilled by processing other data, including synthetic or anonymised data;

+ (b) the special categories of personal data are subject to technical limitations on the re-use of the personal data, and state-of-the-art security and privacy-preserving measures, including pseudonymisation;

+ (c) the special categories of personal data are subject to measures to ensure that the personal data processed are secured, protected, subject to suitable safeguards, including strict controls and documentation of the access, to avoid misuse and ensure that only authorised persons have access to those personal data with appropriate confidentiality obligations;

+ (d) the special categories of personal data are not to be transmitted, transferred or otherwise accessed by other parties;

+ (e) the special categories of personal data are deleted once the bias has been corrected or the personal data has reached the end of its retention period, whichever comes first;

+ (f) the records of processing activities pursuant to Regulations (EU) 2016/679 and (EU) 2018/1725 and Directive (EU) 2016/680 include the reasons why the processing of special categories of personal data was strictly necessary to detect and correct biases, and why that objective could not be achieved by processing other data.

+ 6. For the development of high-risk AI systems not using techniques involving the training of AI models, paragraphs 2 to 5 apply only to the testing data sets.

+ Article 11 Technical documentation

+ 1. The technical documentation of a high-risk AI system shall be drawn up before that system is placed on the market or put into service and shall be kept up-to date.

+ The technical documentation shall be drawn up in such a way as to demonstrate that the high-risk AI system complies with the requirements set out in this Section and to provide national competent authorities and notified bodies with the necessary information in a clear and comprehensive form to assess the compliance of the AI system with those requirements. It shall contain, at a minimum, the elements set out in Annex IV. SMEs, including start-ups, may provide the elements of the technical documentation specified in Annex IV in a simplified manner. To that end, the Commission shall establish a simplified technical documentation form targeted at the needs of small and microenterprises. Where an SME, including a start-up, opts to provide the information required in Annex IV in a simplified manner, it shall use the form referred to in this paragraph. Notified bodies shall accept the form for the purposes of the conformity assessment.

+ 2. Where a high-risk AI system related to a product covered by the Union harmonisation legislation listed in Section A of Annex I is placed on the market or put into service, a single set of technical documentation shall be drawn up containing all the information set out in paragraph 1, as well as the information required under those legal acts.

+ 3. The Commission is empowered to adopt delegated acts in accordance with Article 97 in order to amend Annex IV, where necessary, to ensure that, in light of technical progress, the technical documentation provides all the information necessary to assess the compliance of the system with the requirements set out in this Section.

+ 58/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ Article 12 Record-keeping

+ 1. High-risk AI systems shall technically allow for the automatic recording of events (logs) over the lifetime of the system.

+ 2. In order to ensure a level of traceability of the functioning of a high-risk AI system that is appropriate to the intended purpose of the system, logging capabilities shall enable the recording of events relevant for:

+ (a) identifying situations that may result in the high-risk AI system presenting a risk within the meaning of Article 79(1) or in a substantial modification;

+ (b) facilitating the post-market monitoring referred to in Article 72; and

+ (c) monitoring the operation of high-risk AI systems referred to in Article 26(5).

+ 3. For high-risk AI systems referred to in point 1 (a), of Annex III, the logging capabilities shall provide, at a minimum:

+ (a) recording of the period of each use of the system (start date and time and end date and time of each use);

+ (b) the reference database against which input data has been checked by the system;

+ (c) the input data for which the search has led to a match;

+ (d) the identification of the natural persons involved in the verification of the results, as referred to in Article 14(5).

+ Article 13 Transparency and provision of information to deployers

+ 1. High-risk AI systems shall be designed and developed in such a way as to ensure that their operation is sufficiently transparent to enable deployers to interpret a system's output and use it appropriately. An appropriate type and degree of transparency shall be ensured with a view to achieving compliance with the relevant obligations of the provider and deployer set out in Section 3.

+ 2. High-risk AI systems shall be accompanied by instructions for use in an appropriate digital format or otherwise that include concise, complete, correct and clear information that is relevant, accessible and comprehensible to deployers.

+ 3. The instructions for use shall contain at least the following information:

+ (a) the identity and the contact details of the provider and, where applicable, of its authorised representative;

+ (b) the characteristics, capabilities and limitations of performance of the high-risk AI system, including:

+ (i) its intended purpose;

+ (ii) the level of accuracy, including its metrics, robustness and cybersecurity referred to in Article 15 against which the high-risk AI system has been tested and validated and which can be expected, and any known and foreseeable circumstances that may have an impact on that expected level of accuracy, robustness and cybersecurity;

+ (iii) any known or foreseeable circumstance, related to the use of the high-risk AI system in accordance with its intended purpose or under conditions of reasonably foreseeable misuse, which may lead to risks to the health and safety or fundamental rights referred to in Article 9(2);

+ (iv) where applicable, the technical capabilities and characteristics of the high-risk AI system to provide information that is relevant to explain its output;

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 59/144

+ (v) when appropriate, its performance regarding specific persons or groups of persons on which the system is intended to be used;

+ (vi) when appropriate, specifications for the input data, or any other relevant information in terms of the training, validation and testing data sets used, taking into account the intended purpose of the high-risk AI system;

+ (vii) where applicable, information to enable deployers to interpret the output of the high-risk AI system and use it appropriately;

+ (c) the changes to the high-risk AI system and its performance which have been pre-determined by the provider at the moment of the initial conformity assessment, if any;

+ (d) the human oversight measures referred to in Article 14, including the technical measures put in place to facilitate the interpretation of the outputs of the high-risk AI systems by the deployers;

+ (e) the computational and hardware resources needed, the expected lifetime of the high-risk AI system and any necessary maintenance and care measures, including their frequency, to ensure the proper functioning of that AI system, including as regards software updates;

+ (f) where relevant, a description of the mechanisms included within the high-risk AI system that allows deployers to properly collect, store and interpret the logs in accordance with Article 12.

+ Article 14 Human oversight

+ 1. High-risk AI systems shall be designed and developed in such a way, including with appropriate human-machine interface tools, that they can be effectively overseen by natural persons during the period in which they are in use.

+ 2. Human oversight shall aim to prevent or minimise the risks to health, safety or fundamental rights that may emerge when a high-risk AI system is used in accordance with its intended purpose or under conditions of reasonably foreseeable misuse, in particular where such risks persist despite the application of other requirements set out in this Section.

+ 3. The oversight measures shall be commensurate with the risks, level of autonomy and context of use of the high-risk AI system, and shall be ensured through either one or both of the following types of measures:

+ (a) measures identified and built, when technically feasible, into the high-risk AI system by the provider before it is placed on the market or put into service;

+ (b) measures identified by the provider before placing the high-risk AI system on the market or putting it into service and that are appropriate to be implemented by the deployer.

+ 4. For the purpose of implementing paragraphs 1, 2 and 3, the high-risk AI system shall be provided to the deployer in such a way that natural persons to whom human oversight is assigned are enabled, as appropriate and proportionate:

+ (a) to properly understand the relevant capacities and limitations of the high-risk AI system and be able to duly monitor its operation, including in view of detecting and addressing anomalies, dysfunctions and unexpected performance;

+ (b) to remain aware of the possible tendency of automatically relying or over-relying on the output produced by a high-risk AI system (automation bias), in particular for high-risk AI systems used to provide information or recommendations for decisions to be taken by natural persons;

+ (c) to correctly interpret the high-risk AI system's output, taking into account, for example, the interpretation tools and methods available;

+ 60/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ (d) to decide, in any particular situation, not to use the high-risk AI system or to otherwise disregard, override or reverse the output of the high-risk AI system;

+ (e) to intervene in the operation of the high-risk AI system or interrupt the system through a `stop' button or a similar procedure that allows the system to come to a halt in a safe state.

+ 5. For high-risk AI systems referred to in point 1(a) of Annex III, the measures referred to in paragraph 3 of this Article shall be such as to ensure that, in addition, no action or decision is taken by the deployer on the basis of the identification resulting from the system unless that identification has been separately verified and confirmed by at least two natural persons with the necessary competence, training and authority.

+ The requirement for a separate verification by at least two natural persons shall not apply to high-risk AI systems used for the purposes of law enforcement, migration, border control or asylum, where Union or national law considers the application of this requirement to be disproportionate.

+ Article 15 Accuracy, robustness and cybersecurity

+ 1. High-risk AI systems shall be designed and developed in such a way that they achieve an appropriate level of accuracy, robustness, and cybersecurity, and that they perform consistently in those respects throughout their lifecycle.

+ 2. To address the technical aspects of how to measure the appropriate levels of accuracy and robustness set out in paragraph 1 and any other relevant performance metrics, the Commission shall, in cooperation with relevant stakeholders and organisations such as metrology and benchmarking authorities, encourage, as appropriate, the development of benchmarks and measurement methodologies.

+ 3. The levels of accuracy and the relevant accuracy metrics of high-risk AI systems shall be declared in the accompanying instructions of use.

+ 4. High-risk AI systems shall be as resilient as possible regarding errors, faults or inconsistencies that may occur within the system or the environment in which the system operates, in particular due to their interaction with natural persons or other systems. Technical and organisational measures shall be taken in this regard.

+ The robustness of high-risk AI systems may be achieved through technical redundancy solutions, which may include backup or fail-safe plans.

+ High-risk AI systems that continue to learn after being placed on the market or put into service shall be developed in such a way as to eliminate or reduce as far as possible the risk of possibly biased outputs influencing input for future operations (feedback loops), and as to ensure that any such feedback loops are duly addressed with appropriate mitigation measures.

+ 5. High-risk AI systems shall be resilient against attempts by unauthorised third parties to alter their use, outputs or performance by exploiting system vulnerabilities.

+ The technical solutions aiming to ensure the cybersecurity of high-risk AI systems shall be appropriate to the relevant circumstances and the risks.

+ The technical solutions to address AI specific vulnerabilities shall include, where appropriate, measures to prevent, detect, respond to, resolve and control for attacks trying to manipulate the training data set (data poisoning), or pre-trained components used in training (model poisoning), inputs designed to cause the AI model to make a mistake (adversarial examples or model evasion), confidentiality attacks or model flaws.

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 61/144

+ SECTION 3

+ Obligations of providers and deployers of high-risk AI systems and other parties

+ Article 16 Obligations of providers of high-risk AI systems

+ Providers of high-risk AI systems shall:

+ (a) ensure that their high-risk AI systems are compliant with the requirements set out in Section 2;

+ (b) indicate on the high-risk AI system or, where that is not possible, on its packaging or its accompanying documentation, as applicable, their name, registered trade name or registered trade mark, the address at which they can be contacted;

+ (c) have a quality management system in place which complies with Article 17;

+ (d) keep the documentation referred to in Article 18;

+ (e) when under their control, keep the logs automatically generated by their high-risk AI systems as referred to in Article 19;

+ (f) ensure that the high-risk AI system undergoes the relevant conformity assessment procedure as referred to in Article 43, prior to its being placed on the market or put into service;

+ (g) draw up an EU declaration of conformity in accordance with Article 47;

+ (h) affix the CE marking to the high-risk AI system or, where that is not possible, on its packaging or its accompanying documentation, to indicate conformity with this Regulation, in accordance with Article 48;

+ (i) comply with the registration obligations referred to in Article 49(1);

+ (j) take the necessary corrective actions and provide information as required in Article 20;

+ (k) upon a reasoned request of a national competent authority, demonstrate the conformity of the high-risk AI system with the requirements set out in Section 2;

+ (l) ensure that the high-risk AI system complies with accessibility requirements in accordance with Directives (EU) 2016/2102 and (EU) 2019/882.

+ Article 17 Quality management system

+ 1. Providers of high-risk AI systems shall put a quality management system in place that ensures compliance with this Regulation. That system shall be documented in a systematic and orderly manner in the form of written policies, procedures and instructions, and shall include at least the following aspects:

+ (a) a strategy for regulatory compliance, including compliance with conformity assessment procedures and procedures for the management of modifications to the high-risk AI system;

+ (b) techniques, procedures and systematic actions to be used for the design, design control and design verification of the high-risk AI system;

+ (c) techniques, procedures and systematic actions to be used for the development, quality control and quality assurance of the high-risk AI system;

+ (d) examination, test and validation procedures to be carried out before, during and after the development of the high-risk AI system, and the frequency with which they have to be carried out;

+ 62/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ (e) technical specifications, including standards, to be applied and, where the relevant harmonised standards are not applied in full or do not cover all of the relevant requirements set out in Section 2, the means to be used to ensure that the high-risk AI system complies with those requirements;

+ (f) systems and procedures for data management, including data acquisition, data collection, data analysis, data labelling, data storage, data filtration, data mining, data aggregation, data retention and any other operation regarding the data that is performed before and for the purpose of the placing on the market or the putting into service of high-risk AI systems;

+ (g) the risk management system referred to in Article 9;

+ (h) the setting-up, implementation and maintenance of a post-market monitoring system, in accordance with Article 72;

+ (i) procedures related to the reporting of a serious incident in accordance with Article 73;

+ (j) the handling of communication with national competent authorities, other relevant authorities, including those providing or supporting the access to data, notified bodies, other operators, customers or other interested parties;

+ (k) systems and procedures for record-keeping of all relevant documentation and information;

+ (l) resource management, including security-of-supply related measures;

+ (m) an accountability framework setting out the responsibilities of the management and other staff with regard to all the aspects listed in this paragraph.

+ 2. The implementation of the aspects referred to in paragraph 1 shall be proportionate to the size of the provider's organisation. Providers shall, in any event, respect the degree of rigour and the level of protection required to ensure the compliance of their high-risk AI systems with this Regulation.

+ 3. Providers of high-risk AI systems that are subject to obligations regarding quality management systems or an equivalent function under relevant sectoral Union law may include the aspects listed in paragraph 1 as part of the quality management systems pursuant to that law.

+ 4. For providers that are financial institutions subject to requirements regarding their internal governance, arrangements or processes under Union financial services law, the obligation to put in place a quality management system, with the exception of paragraph 1, points (g), (h) and (i) of this Article, shall be deemed to be fulfilled by complying with the rules on internal governance arrangements or processes pursuant to the relevant Union financial services law. To that end, any harmonised standards referred to in Article 40 shall be taken into account.

+ Article 18 Documentation keeping

+ 1. The provider shall, for a period ending 10 years after the high-risk AI system has been placed on the market or put into service, keep at the disposal of the national competent authorities:

+ (a) the technical documentation referred to in Article 11;

+ (b) the documentation concerning the quality management system referred to in Article 17;

+ (c) the documentation concerning the changes approved by notified bodies, where applicable;

+ (d) the decisions and other documents issued by the notified bodies, where applicable;

+ (e) the EU declaration of conformity referred to in Article 47. ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 63/144

+ 2. Each Member State shall determine conditions under which the documentation referred to in paragraph 1 remains at the disposal of the national competent authorities for the period indicated in that paragraph for the cases when a provider or its authorised representative established on its territory goes bankrupt or ceases its activity prior to the end of that period.

+ 3. Providers that are financial institutions subject to requirements regarding their internal governance, arrangements or processes under Union financial services law shall maintain the technical documentation as part of the documentation kept under the relevant Union financial services law.

+ Article 19 Automatically generated logs

+ 1. Providers of high-risk AI systems shall keep the logs referred to in Article 12(1), automatically generated by their high-risk AI systems, to the extent such logs are under their control. Without prejudice to applicable Union or national law, the logs shall be kept for a period appropriate to the intended purpose of the high-risk AI system, of at least six months, unless provided otherwise in the applicable Union or national law, in particular in Union law on the protection of personal data.

+ 2. Providers that are financial institutions subject to requirements regarding their internal governance, arrangements or processes under Union financial services law shall maintain the logs automatically generated by their high-risk AI systems as part of the documentation kept under the relevant financial services law.

+ Article 20

+ Corrective actions and duty of information

+ 1. Providers of high-risk AI systems which consider or have reason to consider that a high-risk AI system that they have placed on the market or put into service is not in conformity with this Regulation shall immediately take the necessary corrective actions to bring that system into conformity, to withdraw it, to disable it, or to recall it, as appropriate. They shall inform the distributors of the high-risk AI system concerned and, where applicable, the deployers, the authorised representative and importers accordingly.

+ 2. Where the high-risk AI system presents a risk within the meaning of Article 79(1) and the provider becomes aware of that risk, it shall immediately investigate the causes, in collaboration with the reporting deployer, where applicable, and inform the market surveillance authorities competent for the high-risk AI system concerned and, where applicable, the notified body that issued a certificate for that high-risk AI system in accordance with Article 44, in particular, of the nature of the non-compliance and of any relevant corrective action taken.

+ Article 21 Cooperation with competent authorities

+ 1. Providers of high-risk AI systems shall, upon a reasoned request by a competent authority, provide that authority all the information and documentation necessary to demonstrate the conformity of the high-risk AI system with the requirements set out in Section 2, in a language which can be easily understood by the authority in one of the official languages of the institutions of the Union as indicated by the Member State concerned.

+ 2. Upon a reasoned request by a competent authority, providers shall also give the requesting competent authority, as applicable, access to the automatically generated logs of the high-risk AI system referred to in Article 12(1), to the extent such logs are under their control.

+ 3. Any information obtained by a competent authority pursuant to this Article shall be treated in accordance with the confidentiality obligations set out in Article 78.

+ 64/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ Article 22 Authorised representatives of providers of high-risk AI systems

+ 1. Prior to making their high-risk AI systems available on the Union market, providers established in third countries shall, by written mandate, appoint an authorised representative which is established in the Union.

+ 2. The provider shall enable its authorised representative to perform the tasks specified in the mandate received from the provider.

+ 3. The authorised representative shall perform the tasks specified in the mandate received from the provider. It shall provide a copy of the mandate to the market surveillance authorities upon request, in one of the official languages of the institutions of the Union, as indicated by the competent authority. For the purposes of this Regulation, the mandate shall empower the authorised representative to carry out the following tasks:

+ (a) verify that the EU declaration of conformity referred to in Article 47 and the technical documentation referred to in Article 11 have been drawn up and that an appropriate conformity assessment procedure has been carried out by the provider;

+ (b) keep at the disposal of the competent authorities and national authorities or bodies referred to in Article 74(10), for a period of 10 years after the high-risk AI system has been placed on the market or put into service, the contact details of the provider that appointed the authorised representative, a copy of the EU declaration of conformity referred to in Article 47, the technical documentation and, if applicable, the certificate issued by the notified body;

+ (c) provide a competent authority, upon a reasoned request, with all the information and documentation, including that referred to in point (b) of this subparagraph, necessary to demonstrate the conformity of a high-risk AI system with the requirements set out in Section 2, including access to the logs, as referred to in Article 12(1), automatically generated by the high-risk AI system, to the extent such logs are under the control of the provider;

+ (d) cooperate with competent authorities, upon a reasoned request, in any action the latter take in relation to the high-risk AI system, in particular to reduce and mitigate the risks posed by the high-risk AI system;

+ (e) where applicable, comply with the registration obligations referred to in Article 49(1), or, if the registration is carried out by the provider itself, ensure that the information referred to in point 3 of Section A of Annex VIII is correct.

+ The mandate shall empower the authorised representative to be addressed, in addition to or instead of the provider, by the competent authorities, on all issues related to ensuring compliance with this Regulation.

+ 4. The authorised representative shall terminate the mandate if it considers or has reason to consider the provider to be acting contrary to its obligations pursuant to this Regulation. In such a case, it shall immediately inform the relevant market surveillance authority, as well as, where applicable, the relevant notified body, about the termination of the mandate and the reasons therefor.

+ Article 23 Obligations of importers

+ 1. Before placing a high-risk AI system on the market, importers shall ensure that the system is in conformity with this Regulation by verifying that:

+ (a) the relevant conformity assessment procedure referred to in Article 43 has been carried out by the provider of the high-risk AI system;

+ (b) the provider has drawn up the technical documentation in accordance with Article 11 and Annex IV;

+ (c) the system bears the required CE marking and is accompanied by the EU declaration of conformity referred to in Article 47 and instructions for use;

+ (d) the provider has appointed an authorised representative in accordance with Article 22(1). ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 65/144

+ 2. Where an importer has sufficient reason to consider that a high-risk AI system is not in conformity with this Regulation, or is falsified, or accompanied by falsified documentation, it shall not place the system on the market until it has been brought into conformity. Where the high-risk AI system presents a risk within the meaning of Article 79(1), the importer shall inform the provider of the system, the authorised representative and the market surveillance authorities to that effect.

+ 3. Importers shall indicate their name, registered trade name or registered trade mark, and the address at which they can be contacted on the high-risk AI system and on its packaging or its accompanying documentation, where applicable.

+ 4. Importers shall ensure that, while a high-risk AI system is under their responsibility, storage or transport conditions, where applicable, do not jeopardise its compliance with the requirements set out in Section 2.

+ 5. Importers shall keep, for a period of 10 years after the high-risk AI system has been placed on the market or put into service, a copy of the certificate issued by the notified body, where applicable, of the instructions for use, and of the EU declaration of conformity referred to in Article 47.

+ 6. Importers shall provide the relevant competent authorities, upon a reasoned request, with all the necessary information and documentation, including that referred to in paragraph 5, to demonstrate the conformity of a high-risk AI system with the requirements set out in Section 2 in a language which can be easily understood by them. For this purpose, they shall also ensure that the technical documentation can be made available to those authorities.

+ 7. Importers shall cooperate with the relevant competent authorities in any action those authorities take in relation to a high-risk AI system placed on the market by the importers, in particular to reduce and mitigate the risks posed by it.

+ Article 24 Obligations of distributors

+ 1. Before making a high-risk AI system available on the market, distributors shall verify that it bears the required CE marking, that it is accompanied by a copy of the EU declaration of conformity referred to in Article 47 and instructions for use, and that the provider and the importer of that system, as applicable, have complied with their respective obligations as laid down in Article 16, points (b) and (c) and Article 23(3).

+ 2. Where a distributor considers or has reason to consider, on the basis of the information in its possession, that a high-risk AI system is not in conformity with the requirements set out in Section 2, it shall not make the high-risk AI system available on the market until the system has been brought into conformity with those requirements. Furthermore, where the high-risk AI system presents a risk within the meaning of Article 79(1), the distributor shall inform the provider or the importer of the system, as applicable, to that effect.

+ 3. Distributors shall ensure that, while a high-risk AI system is under their responsibility, storage or transport conditions, where applicable, do not jeopardise the compliance of the system with the requirements set out in Section 2.

+ 4. A distributor that considers or has reason to consider, on the basis of the information in its possession, a high-risk AI system which it has made available on the market not to be in conformity with the requirements set out in Section 2, shall take the corrective actions necessary to bring that system into conformity with those requirements, to withdraw it or recall it, or shall ensure that the provider, the importer or any relevant operator, as appropriate, takes those corrective actions. Where the high-risk AI system presents a risk within the meaning of Article 79(1), the distributor shall immediately inform the provider or importer of the system and the authorities competent for the high-risk AI system concerned, giving details, in particular, of the non-compliance and of any corrective actions taken.

+ 5. Upon a reasoned request from a relevant competent authority, distributors of a high-risk AI system shall provide that authority with all the information and documentation regarding their actions pursuant to paragraphs 1 to 4 necessary to demonstrate the conformity of that system with the requirements set out in Section 2.

+ 6. Distributors shall cooperate with the relevant competent authorities in any action those authorities take in relation to a high-risk AI system made available on the market by the distributors, in particular to reduce or mitigate the risk posed by it.

+ 66/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ Article 25 Responsibilities along the AI value chain

+ 1. Any distributor, importer, deployer or other third-party shall be considered to be a provider of a high-risk AI system for the purposes of this Regulation and shall be subject to the obligations of the provider under Article 16, in any of the following circumstances:

+ (a) they put their name or trademark on a high-risk AI system already placed on the market or put into service, without prejudice to contractual arrangements stipulating that the obligations are otherwise allocated;

+ (b) they make a substantial modification to a high-risk AI system that has already been placed on the market or has already been put into service in such a way that it remains a high-risk AI system pursuant to Article 6;

+ (c) they modify the intended purpose of an AI system, including a general-purpose AI system, which has not been classified as high-risk and has already been placed on the market or put into service in such a way that the AI system concerned becomes a high-risk AI system in accordance with Article 6.

+ 2. Where the circumstances referred to in paragraph 1 occur, the provider that initially placed the AI system on the market or put it into service shall no longer be considered to be a provider of that specific AI system for the purposes of this Regulation. That initial provider shall closely cooperate with new providers and shall make available the necessary information and provide the reasonably expected technical access and other assistance that are required for the fulfilment of the obligations set out in this Regulation, in particular regarding the compliance with the conformity assessment of high-risk AI systems. This paragraph shall not apply in cases where the initial provider has clearly specified that its AI system is not to be changed into a high-risk AI system and therefore does not fall under the obligation to hand over the documentation.

+ 3. In the case of high-risk AI systems that are safety components of products covered by the Union harmonisation legislation listed in Section A of Annex I, the product manufacturer shall be considered to be the provider of the high-risk AI system, and shall be subject to the obligations under Article 16 under either of the following circumstances:

+ (a) the high-risk AI system is placed on the market together with the product under the name or trademark of the product manufacturer;

+ (b) the high-risk AI system is put into service under the name or trademark of the product manufacturer after the product has been placed on the market.

+ 4. The provider of a high-risk AI system and the third party that supplies an AI system, tools, services, components, or processes that are used or integrated in a high-risk AI system shall, by written agreement, specify the necessary information, capabilities, technical access and other assistance based on the generally acknowledged state of the art, in order to enable the provider of the high-risk AI system to fully comply with the obligations set out in this Regulation. This paragraph shall not apply to third parties making accessible to the public tools, services, processes, or components, other than general-purpose AI models, under a free and open-source licence.

+ The AI Office may develop and recommend voluntary model terms for contracts between providers of high-risk AI systems and third parties that supply tools, services, components or processes that are used for or integrated into high-risk AI systems. When developing those voluntary model terms, the AI Office shall take into account possible contractual requirements applicable in specific sectors or business cases. The voluntary model terms shall be published and be available free of charge in an easily usable electronic format.

+ 5. Paragraphs 2 and 3 are without prejudice to the need to observe and protect intellectual property rights, confidential business information and trade secrets in accordance with Union and national law.

+ Article 26 Obligations of deployers of high-risk AI systems

+ 1. Deployers of high-risk AI systems shall take appropriate technical and organisational measures to ensure they use such systems in accordance with the instructions for use accompanying the systems, pursuant to paragraphs 3 and 6.

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 67/144

+ 2. Deployers shall assign human oversight to natural persons who have the necessary competence, training and authority, as well as the necessary support.

+ 3. The obligations set out in paragraphs 1 and 2, are without prejudice to other deployer obligations under Union or national law and to the deployer's freedom to organise its own resources and activities for the purpose of implementing the human oversight measures indicated by the provider.

+ 4. Without prejudice to paragraphs 1 and 2, to the extent the deployer exercises control over the input data, that deployer shall ensure that input data is relevant and sufficiently representative in view of the intended purpose of the high-risk AI system.

+ 5. Deployers shall monitor the operation of the high-risk AI system on the basis of the instructions for use and, where relevant, inform providers in accordance with Article 72. Where deployers have reason to consider that the use of the high-risk AI system in accordance with the instructions may result in that AI system presenting a risk within the meaning of Article 79(1), they shall, without undue delay, inform the provider or distributor and the relevant market surveillance authority, and shall suspend the use of that system. Where deployers have identified a serious incident, they shall also immediately inform first the provider, and then the importer or distributor and the relevant market surveillance authorities of that incident. If the deployer is not able to reach the provider, Article 73 shall apply mutatis mutandis. This obligation shall not cover sensitive operational data of deployers of AI systems which are law enforcement authorities.

+ For deployers that are financial institutions subject to requirements regarding their internal governance, arrangements or processes under Union financial services law, the monitoring obligation set out in the first subparagraph shall be deemed to be fulfilled by complying with the rules on internal governance arrangements, processes and mechanisms pursuant to the relevant financial service law.

+ 6. Deployers of high-risk AI systems shall keep the logs automatically generated by that high-risk AI system to the extent such logs are under their control, for a period appropriate to the intended purpose of the high-risk AI system, of at least six months, unless provided otherwise in applicable Union or national law, in particular in Union law on the protection of personal data.

+ Deployers that are financial institutions subject to requirements regarding their internal governance, arrangements or processes under Union financial services law shall maintain the logs as part of the documentation kept pursuant to the relevant Union financial service law.

+ 7. Before putting into service or using a high-risk AI system at the workplace, deployers who are employers shall inform workers' representatives and the affected workers that they will be subject to the use of the high-risk AI system. This information shall be provided, where applicable, in accordance with the rules and procedures laid down in Union and national law and practice on information of workers and their representatives.

+ 8. Deployers of high-risk AI systems that are public authorities, or Union institutions, bodies, offices or agencies shall comply with the registration obligations referred to in Article 49. When such deployers find that the high-risk AI system that they envisage using has not been registered in the EU database referred to in Article 71, they shall not use that system and shall inform the provider or the distributor.

+ 9. Where applicable, deployers of high-risk AI systems shall use the information provided under Article 13 of this Regulation to comply with their obligation to carry out a data protection impact assessment under Article 35 of Regulation (EU) 2016/679 or Article 27 of Directive (EU) 2016/680.

+ 10. Without prejudice to Directive (EU) 2016/680, in the framework of an investigation for the targeted search of a person suspected or convicted of having committed a criminal offence, the deployer of a high-risk AI system for post-remote biometric identification shall request an authorisation, ex ante, or without undue delay and no later than 48 hours, by a judicial authority or an administrative authority whose decision is binding and subject to judicial review, for the use of that system, except when it is used for the initial identification of a potential suspect based on objective and verifiable facts directly linked to the offence. Each use shall be limited to what is strictly necessary for the investigation of a specific criminal offence.

+ If the authorisation requested pursuant to the first subparagraph is rejected, the use of the post-remote biometric identification system linked to that requested authorisation shall be stopped with immediate effect and the personal data linked to the use of the high-risk AI system for which the authorisation was requested shall be deleted.

+ 68/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ In no case shall such high-risk AI system for post-remote biometric identification be used for law enforcement purposes in an untargeted way, without any link to a criminal offence, a criminal proceeding, a genuine and present or genuine and foreseeable threat of a criminal offence, or the search for a specific missing person. It shall be ensured that no decision that produces an adverse legal effect on a person may be taken by the law enforcement authorities based solely on the output of such post-remote biometric identification systems.

+ This paragraph is without prejudice to Article 9 of Regulation (EU) 2016/679 and Article 10 of Directive (EU) 2016/680 for the processing of biometric data.

+ Regardless of the purpose or deployer, each use of such high-risk AI systems shall be documented in the relevant police file and shall be made available to the relevant market surveillance authority and the national data protection authority upon request, excluding the disclosure of sensitive operational data related to law enforcement. This subparagraph shall be without prejudice to the powers conferred by Directive (EU) 2016/680 on supervisory authorities.

+ Deployers shall submit annual reports to the relevant market surveillance and national data protection authorities on their use of post-remote biometric identification systems, excluding the disclosure of sensitive operational data related to law enforcement. The reports may be aggregated to cover more than one deployment.

+ Member States may introduce, in accordance with Union law, more restrictive laws on the use of post-remote biometric identification systems.

+ 11. Without prejudice to Article 50 of this Regulation, deployers of high-risk AI systems referred to in Annex III that make decisions or assist in making decisions related to natural persons shall inform the natural persons that they are subject to the use of the high-risk AI system. For high-risk AI systems used for law enforcement purposes Article 13 of Directive (EU) 2016/680 shall apply.

+ 12. Deployers shall cooperate with the relevant competent authorities in any action those authorities take in relation to the high-risk AI system in order to implement this Regulation.

+ Article 27 Fundamental rights impact assessment for high-risk AI systems

+ 1. Prior to deploying a high-risk AI system referred to in Article 6(2), with the exception of high-risk AI systems intended to be used in the area listed in point 2 of Annex III, deployers that are bodies governed by public law, or are private entities providing public services, and deployers of high-risk AI systems referred to in points 5 (b) and (c) of Annex III, shall perform an assessment of the impact on fundamental rights that the use of such system may produce. For that purpose, deployers shall perform an assessment consisting of:

+ (a) a description of the deployer's processes in which the high-risk AI system will be used in line with its intended purpose;

+ (b) a description of the period of time within which, and the frequency with which, each high-risk AI system is intended to be used;

+ (c) the categories of natural persons and groups likely to be affected by its use in the specific context;

+ (d) the specific risks of harm likely to have an impact on the categories of natural persons or groups of persons identified pursuant to point (c) of this paragraph, taking into account the information given by the provider pursuant to Article 13;

+ (e) a description of the implementation of human oversight measures, according to the instructions for use;

+ (f) the measures to be taken in the case of the materialisation of those risks, including the arrangements for internal governance and complaint mechanisms.

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 69/144

+ 2. The obligation laid down in paragraph 1 applies to the first use of the high-risk AI system. The deployer may, in similar cases, rely on previously conducted fundamental rights impact assessments or existing impact assessments carried out by provider. If, during the use of the high-risk AI system, the deployer considers that any of the elements listed in paragraph 1 has changed or is no longer up to date, the deployer shall take the necessary steps to update the information.

+ 3. Once the assessment referred to in paragraph 1 of this Article has been performed, the deployer shall notify the market surveillance authority of its results, submitting the filled-out template referred to in paragraph 5 of this Article as part of the notification. In the case referred to in Article 46(1), deployers may be exempt from that obligation to notify.

+ 4. If any of the obligations laid down in this Article is already met through the data protection impact assessment conducted pursuant to Article 35 of Regulation (EU) 2016/679 or Article 27 of Directive (EU) 2016/680, the fundamental rights impact assessment referred to in paragraph 1 of this Article shall complement that data protection impact assessment.

+ 5. The AI Office shall develop a template for a questionnaire, including through an automated tool, to facilitate deployers in complying with their obligations under this Article in a simplified manner.

+ SECTION 4

+ Notifying authorities and notified bodies

+ Article 28 Notifying authorities

+ 1. Each Member State shall designate or establish at least one notifying authority responsible for setting up and carrying out the necessary procedures for the assessment, designation and notification of conformity assessment bodies and for their monitoring. Those procedures shall be developed in cooperation between the notifying authorities of all Member States.

+ 2. Member States may decide that the assessment and monitoring referred to in paragraph 1 is to be carried out by a national accreditation body within the meaning of, and in accordance with, Regulation (EC) No 765/2008.

+ 3. Notifying authorities shall be established, organised and operated in such a way that no conflict of interest arises with conformity assessment bodies, and that the objectivity and impartiality of their activities are safeguarded.

+ 4. Notifying authorities shall be organised in such a way that decisions relating to the notification of conformity assessment bodies are taken by competent persons different from those who carried out the assessment of those bodies.

+ 5. Notifying authorities shall offer or provide neither any activities that conformity assessment bodies perform, nor any consultancy services on a commercial or competitive basis.

+ 6. Notifying authorities shall safeguard the confidentiality of the information that they obtain, in accordance with Article 78.

+ 7. Notifying authorities shall have an adequate number of competent personnel at their disposal for the proper performance of their tasks. Competent personnel shall have the necessary expertise, where applicable, for their function, in fields such as information technologies, AI and law, including the supervision of fundamental rights.

+ Article 29 Application of a conformity assessment body for notification

+ 1. Conformity assessment bodies shall submit an application for notification to the notifying authority of the Member State in which they are established.

+ 70/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 2. The application for notification shall be accompanied by a description of the conformity assessment activities, the conformity assessment module or modules and the types of AI systems for which the conformity assessment body claims to be competent, as well as by an accreditation certificate, where one exists, issued by a national accreditation body attesting that the conformity assessment body fulfils the requirements laid down in Article 31.

+ Any valid document related to existing designations of the applicant notified body under any other Union harmonisation legislation shall be added.

+ 3. Where the conformity assessment body concerned cannot provide an accreditation certificate, it shall provide the notifying authority with all the documentary evidence necessary for the verification, recognition and regular monitoring of its compliance with the requirements laid down in Article 31.

+ 4. For notified bodies which are designated under any other Union harmonisation legislation, all documents and certificates linked to those designations may be used to support their designation procedure under this Regulation, as appropriate. The notified body shall update the documentation referred to in paragraphs 2 and 3 of this Article whenever relevant changes occur, in order to enable the authority responsible for notified bodies to monitor and verify continuous compliance with all the requirements laid down in Article 31.

+ Article 30 Notification procedure

+ 1. Notifying authorities may notify only conformity assessment bodies which have satisfied the requirements laid down in Article 31.

+ 2. Notifying authorities shall notify the Commission and the other Member States, using the electronic notification tool developed and managed by the Commission, of each conformity assessment body referred to in paragraph 1.

+ 3. The notification referred to in paragraph 2 of this Article shall include full details of the conformity assessment activities, the conformity assessment module or modules, the types of AI systems concerned, and the relevant attestation of competence. Where a notification is not based on an accreditation certificate as referred to in Article 29(2), the notifying authority shall provide the Commission and the other Member States with documentary evidence which attests to the competence of the conformity assessment body and to the arrangements in place to ensure that that body will be monitored regularly and will continue to satisfy the requirements laid down in Article 31.

+ 4. The conformity assessment body concerned may perform the activities of a notified body only where no objections are raised by the Commission or the other Member States within two weeks of a notification by a notifying authority where it includes an accreditation certificate referred to in Article 29(2), or within two months of a notification by the notifying authority where it includes documentary evidence referred to in Article 29(3).

+ 5. Where objections are raised, the Commission shall, without delay, enter into consultations with the relevant Member States and the conformity assessment body. In view thereof, the Commission shall decide whether the authorisation is justified. The Commission shall address its decision to the Member State concerned and to the relevant conformity assessment body.

+ Article 31 Requirements relating to notified bodies

+ 1. A notified body shall be established under the national law of a Member State and shall have legal personality.

+ 2. Notified bodies shall satisfy the organisational, quality management, resources and process requirements that are necessary to fulfil their tasks, as well as suitable cybersecurity requirements.

+ 3. The organisational structure, allocation of responsibilities, reporting lines and operation of notified bodies shall ensure confidence in their performance, and in the results of the conformity assessment activities that the notified bodies conduct.

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 71/144

+ 4. Notified bodies shall be independent of the provider of a high-risk AI system in relation to which they perform conformity assessment activities. Notified bodies shall also be independent of any other operator having an economic interest in high-risk AI systems assessed, as well as of any competitors of the provider. This shall not preclude the use of assessed high-risk AI systems that are necessary for the operations of the conformity assessment body, or the use of such high-risk AI systems for personal purposes.

+ 5. Neither a conformity assessment body, its top-level management nor the personnel responsible for carrying out its conformity assessment tasks shall be directly involved in the design, development, marketing or use of high-risk AI systems, nor shall they represent the parties engaged in those activities. They shall not engage in any activity that might conflict with their independence of judgement or integrity in relation to conformity assessment activities for which they are notified. This shall, in particular, apply to consultancy services.

+ 6. Notified bodies shall be organised and operated so as to safeguard the independence, objectivity and impartiality of their activities. Notified bodies shall document and implement a structure and procedures to safeguard impartiality and to promote and apply the principles of impartiality throughout their organisation, personnel and assessment activities.

+ 7. Notified bodies shall have documented procedures in place ensuring that their personnel, committees, subsidiaries, subcontractors and any associated body or personnel of external bodies maintain, in accordance with Article 78, the confidentiality of the information which comes into their possession during the performance of conformity assessment activities, except when its disclosure is required by law. The staff of notified bodies shall be bound to observe professional secrecy with regard to all information obtained in carrying out their tasks under this Regulation, except in relation to the notifying authorities of the Member State in which their activities are carried out.

+ 8. Notified bodies shall have procedures for the performance of activities which take due account of the size of a provider, the sector in which it operates, its structure, and the degree of complexity of the AI system concerned.

+ 9. Notified bodies shall take out appropriate liability insurance for their conformity assessment activities, unless liability is assumed by the Member State in which they are established in accordance with national law or that Member State is itself directly responsible for the conformity assessment.

+ 10. Notified bodies shall be capable of carrying out all their tasks under this Regulation with the highest degree of professional integrity and the requisite competence in the specific field, whether those tasks are carried out by notified bodies themselves or on their behalf and under their responsibility.

+ 11. Notified bodies shall have sufficient internal competences to be able effectively to evaluate the tasks conducted by external parties on their behalf. The notified body shall have permanent availability of sufficient administrative, technical, legal and scientific personnel who possess experience and knowledge relating to the relevant types of AI systems, data and data computing, and relating to the requirements set out in Section 2.

+ 12. Notified bodies shall participate in coordination activities as referred to in Article 38. They shall also take part directly, or be represented in, European standardisation organisations, or ensure that they are aware and up to date in respect of relevant standards.

+ Article 32 Presumption of conformity with requirements relating to notified bodies

+ Where a conformity assessment body demonstrates its conformity with the criteria laid down in the relevant harmonised standards or parts thereof, the references of which have been published in the Official Journal of the European Union, it shall be presumed to comply with the requirements set out in Article 31 in so far as the applicable harmonised standards cover those requirements.

+ 72/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ Article 33 Subsidiaries of notified bodies and subcontracting

+ 1. Where a notified body subcontracts specific tasks connected with the conformity assessment or has recourse to a subsidiary, it shall ensure that the subcontractor or the subsidiary meets the requirements laid down in Article 31, and shall inform the notifying authority accordingly.

+ 2. Notified bodies shall take full responsibility for the tasks performed by any subcontractors or subsidiaries.

+ 3. Activities may be subcontracted or carried out by a subsidiary only with the agreement of the provider. Notified bodies shall make a list of their subsidiaries publicly available.

+ 4. The relevant documents concerning the assessment of the qualifications of the subcontractor or the subsidiary and the work carried out by them under this Regulation shall be kept at the disposal of the notifying authority for a period of five years from the termination date of the subcontracting.

+ Article 34 Operational obligations of notified bodies

+ 1. Notified bodies shall verify the conformity of high-risk AI systems in accordance with the conformity assessment procedures set out in Article 43.

+ 2. Notified bodies shall avoid unnecessary burdens for providers when performing their activities, and take due account of the size of the provider, the sector in which it operates, its structure and the degree of complexity of the high-risk AI system concerned, in particular in view of minimising administrative burdens and compliance costs for micro- and small enterprises within the meaning of Recommendation 2003/361/EC. The notified body shall, nevertheless, respect the degree of rigour and the level of protection required for the compliance of the high-risk AI system with the requirements of this Regulation.

+ 3. Notified bodies shall make available and submit upon request all relevant documentation, including the providers' documentation, to the notifying authority referred to in Article 28 to allow that authority to conduct its assessment, designation, notification and monitoring activities, and to facilitate the assessment outlined in this Section.

+ Article 35 Identification numbers and lists of notified bodies

+ 1. The Commission shall assign a single identification number to each notified body, even where a body is notified under more than one Union act.

+ 2. The Commission shall make publicly available the list of the bodies notified under this Regulation, including their identification numbers and the activities for which they have been notified. The Commission shall ensure that the list is kept up to date.

+ Article 36 Changes to notifications

+ 1. The notifying authority shall notify the Commission and the other Member States of any relevant changes to the notification of a notified body via the electronic notification tool referred to in Article 30(2).

+ 2. The procedures laid down in Articles 29 and 30 shall apply to extensions of the scope of the notification.

+ For changes to the notification other than extensions of its scope, the procedures laid down in paragraphs (3) to (9) shall apply.

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 73/144

+ 3. Where a notified body decides to cease its conformity assessment activities, it shall inform the notifying authority and the providers concerned as soon as possible and, in the case of a planned cessation, at least one year before ceasing its activities. The certificates of the notified body may remain valid for a period of nine months after cessation of the notified body's activities, on condition that another notified body has confirmed in writing that it will assume responsibilities for the high-risk AI systems covered by those certificates. The latter notified body shall complete a full assessment of the high-risk AI systems affected by the end of that nine-month-period before issuing new certificates for those systems. Where the notified body has ceased its activity, the notifying authority shall withdraw the designation.

+ 4. Where a notifying authority has sufficient reason to consider that a notified body no longer meets the requirements laid down in Article 31, or that it is failing to fulfil its obligations, the notifying authority shall without delay investigate the matter with the utmost diligence. In that context, it shall inform the notified body concerned about the objections raised and give it the possibility to make its views known. If the notifying authority comes to the conclusion that the notified body no longer meets the requirements laid down in Article 31 or that it is failing to fulfil its obligations, it shall restrict, suspend or withdraw the designation as appropriate, depending on the seriousness of the failure to meet those requirements or fulfil those obligations. It shall immediately inform the Commission and the other Member States accordingly.

+ 5. Where its designation has been suspended, restricted, or fully or partially withdrawn, the notified body shall inform the providers concerned within 10 days.

+ 6. In the event of the restriction, suspension or withdrawal of a designation, the notifying authority shall take appropriate steps to ensure that the files of the notified body concerned are kept, and to make them available to notifying authorities in other Member States and to market surveillance authorities at their request.

+ 7. In the event of the restriction, suspension or withdrawal of a designation, the notifying authority shall:

+ (a) assess the impact on the certificates issued by the notified body;

+ (b) submit a report on its findings to the Commission and the other Member States within three months of having notified the changes to the designation;

+ (c) require the notified body to suspend or withdraw, within a reasonable period of time determined by the authority, any certificates which were unduly issued, in order to ensure the continuing conformity of high-risk AI systems on the market;

+ (d) inform the Commission and the Member States about certificates the suspension or withdrawal of which it has required;

+ (e) provide the national competent authorities of the Member State in which the provider has its registered place of business with all relevant information about the certificates of which it has required the suspension or withdrawal; that authority shall take the appropriate measures, where necessary, to avoid a potential risk to health, safety or fundamental rights.

+ 8. With the exception of certificates unduly issued, and where a designation has been suspended or restricted, the certificates shall remain valid in one of the following circumstances:

+ (a) the notifying authority has confirmed, within one month of the suspension or restriction, that there is no risk to health, safety or fundamental rights in relation to certificates affected by the suspension or restriction, and the notifying authority has outlined a timeline for actions to remedy the suspension or restriction; or

+ (b) the notifying authority has confirmed that no certificates relevant to the suspension will be issued, amended or re-issued during the course of the suspension or restriction, and states whether the notified body has the capability of continuing to monitor and remain responsible for existing certificates issued for the period of the suspension or restriction; in the event that the notifying authority determines that the notified body does not have the capability to support existing certificates issued, the provider of the system covered by the certificate shall confirm in writing to the national competent authorities of the Member State in which it has its registered place of business, within three months of the suspension or restriction, that another qualified notified body is temporarily assuming the functions of the notified body to monitor and remain responsible for the certificates during the period of suspension or restriction.

+ 74/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 9. With the exception of certificates unduly issued, and where a designation has been withdrawn, the certificates shall remain valid for a period of nine months under the following circumstances:

+ (a) the national competent authority of the Member State in which the provider of the high-risk AI system covered by the certificate has its registered place of business has confirmed that there is no risk to health, safety or fundamental rights associated with the high-risk AI systems concerned; and

+ (b) another notified body has confirmed in writing that it will assume immediate responsibility for those AI systems and completes its assessment within 12 months of the withdrawal of the designation.

+ In the circumstances referred to in the first subparagraph, the national competent authority of the Member State in which the provider of the system covered by the certificate has its place of business may extend the provisional validity of the certificates for additional periods of three months, which shall not exceed 12 months in total.

+ The national competent authority or the notified body assuming the functions of the notified body affected by the change of designation shall immediately inform the Commission, the other Member States and the other notified bodies thereof.

+ Article 37 Challenge to the competence of notified bodies

+ 1. The Commission shall, where necessary, investigate all cases where there are reasons to doubt the competence of a notified body or the continued fulfilment by a notified body of the requirements laid down in Article 31 and of its applicable responsibilities.

+ 2. The notifying authority shall provide the Commission, on request, with all relevant information relating to the notification or the maintenance of the competence of the notified body concerned.

+ 3. The Commission shall ensure that all sensitive information obtained in the course of its investigations pursuant to this Article is treated confidentially in accordance with Article 78.

+ 4. Where the Commission ascertains that a notified body does not meet or no longer meets the requirements for its notification, it shall inform the notifying Member State accordingly and request it to take the necessary corrective measures, including the suspension or withdrawal of the notification if necessary. Where the Member State fails to take the necessary corrective measures, the Commission may, by means of an implementing act, suspend, restrict or withdraw the designation. That implementing act shall be adopted in accordance with the examination procedure referred to in Article 98(2).

+ Article 38 Coordination of notified bodies

+ 1. The Commission shall ensure that, with regard to high-risk AI systems, appropriate coordination and cooperation between notified bodies active in the conformity assessment procedures pursuant to this Regulation are put in place and properly operated in the form of a sectoral group of notified bodies.

+ 2. Each notifying authority shall ensure that the bodies notified by it participate in the work of a group referred to in paragraph 1, directly or through designated representatives.

+ 3. The Commission shall provide for the exchange of knowledge and best practices between notifying authorities.

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 75/144

+ Article 39 Conformity assessment bodies of third countries

+ Conformity assessment bodies established under the law of a third country with which the Union has concluded an agreement may be authorised to carry out the activities of notified bodies under this Regulation, provided that they meet the requirements laid down in Article 31 or they ensure an equivalent level of compliance.

+ SECTION 5

+ Standards, conformity assessment, certificates, registration

+ Article 40

+ Harmonised standards and standardisation deliverables

+ 1. High-risk AI systems or general-purpose AI models which are in conformity with harmonised standards or parts thereof the references of which have been published in the Official Journal of the European Union in accordance with Regulation (EU) No 1025/2012 shall be presumed to be in conformity with the requirements set out in Section 2 of this Chapter or, as applicable, with the obligations set out in of Chapter V, Sections 2 and 3, of this Regulation, to the extent that those standards cover those requirements or obligations.

+ 2. In accordance with Article 10 of Regulation (EU) No 1025/2012, the Commission shall issue, without undue delay, standardisation requests covering all requirements set out in Section 2 of this Chapter and, as applicable, standardisation requests covering obligations set out in Chapter V, Sections 2 and 3, of this Regulation. The standardisation request shall also ask for deliverables on reporting and documentation processes to improve AI systems' resource performance, such as reducing the high-risk AI system's consumption of energy and of other resources during its lifecycle, and on the energy-efficient development of general-purpose AI models. When preparing a standardisation request, the Commission shall consult the Board and relevant stakeholders, including the advisory forum.

+ When issuing a standardisation request to European standardisation organisations, the Commission shall specify that standards have to be clear, consistent, including with the standards developed in the various sectors for products covered by the existing Union harmonisation legislation listed in Annex I, and aiming to ensure that high-risk AI systems or general-purpose AI models placed on the market or put into service in the Union meet the relevant requirements or obligations laid down in this Regulation.

+ The Commission shall request the European standardisation organisations to provide evidence of their best efforts to fulfil the objectives referred to in the first and the second subparagraph of this paragraph in accordance with Article 24 of Regulation (EU) No 1025/2012.

+ 3. The participants in the standardisation process shall seek to promote investment and innovation in AI, including through increasing legal certainty, as well as the competitiveness and growth of the Union market, to contribute to strengthening global cooperation on standardisation and taking into account existing international standards in the field of AI that are consistent with Union values, fundamental rights and interests, and to enhance multi-stakeholder governance ensuring a balanced representation of interests and the effective participation of all relevant stakeholders in accordance with Articles 5, 6, and 7 of Regulation (EU) No 1025/2012.

+ Article 41 Common specifications

+ 1. The Commission may adopt, implementing acts establishing common specifications for the requirements set out in Section 2 of this Chapter or, as applicable, for the obligations set out in Sections 2 and 3 of Chapter V where the following conditions have been fulfilled:

+ (a) the Commission has requested, pursuant to Article 10(1) of Regulation (EU) No 1025/2012, one or more European standardisation organisations to draft a harmonised standard for the requirements set out in Section 2 of this Chapter, or, as applicable, for the obligations set out in Sections 2 and 3 of Chapter V, and:

+ (i) the request has not been accepted by any of the European standardisation organisations; or

+ 76/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ (ii) the harmonised standards addressing that request are not delivered within the deadline set in accordance with Article 10(1) of Regulation (EU) No 1025/2012; or

+ (iii) the relevant harmonised standards insufficiently address fundamental rights concerns; or

+ (iv) the harmonised standards do not comply with the request; and

+ (b) no reference to harmonised standards covering the requirements referred to in Section 2 of this Chapter or, as applicable, the obligations referred to in Sections 2 and 3 of Chapter V has been published in the Official Journal of the European Union in accordance with Regulation (EU) No 1025/2012, and no such reference is expected to be published within a reasonable period.

+ When drafting the common specifications, the Commission shall consult the advisory forum referred to in Article 67.

+ The implementing acts referred to in the first subparagraph of this paragraph shall be adopted in accordance with the examination procedure referred to in Article 98(2).

+ 2. Before preparing a draft implementing act, the Commission shall inform the committee referred to in Article 22 of Regulation (EU) No 1025/2012 that it considers the conditions laid down in paragraph 1 of this Article to be fulfilled.

+ 3. High-risk AI systems or general-purpose AI models which are in conformity with the common specifications referred to in paragraph 1, or parts of those specifications, shall be presumed to be in conformity with the requirements set out in Section 2 of this Chapter or, as applicable, to comply with the obligations referred to in Sections 2 and 3 of Chapter V, to the extent those common specifications cover those requirements or those obligations.

+ 4. Where a harmonised standard is adopted by a European standardisation organisation and proposed to the Commission for the publication of its reference in the Official Journal of the European Union, the Commission shall assess the harmonised standard in accordance with Regulation (EU) No 1025/2012. When reference to a harmonised standard is published in the Official Journal of the European Union, the Commission shall repeal the implementing acts referred to in paragraph 1, or parts thereof which cover the same requirements set out in Section 2 of this Chapter or, as applicable, the same obligations set out in Sections 2 and 3 of Chapter V.

+ 5. Where providers of high-risk AI systems or general-purpose AI models do not comply with the common specifications referred to in paragraph 1, they shall duly justify that they have adopted technical solutions that meet the requirements referred to in Section 2 of this Chapter or, as applicable, comply with the obligations set out in Sections 2 and 3 of Chapter V to a level at least equivalent thereto.

+ 6. Where a Member State considers that a common specification does not entirely meet the requirements set out in Section 2 or, as applicable, comply with obligations set out in Sections 2 and 3 of Chapter V, it shall inform the Commission thereof with a detailed explanation. The Commission shall assess that information and, if appropriate, amend the implementing act establishing the common specification concerned.

+ Article 42 Presumption of conformity with certain requirements

+ 1. High-risk AI systems that have been trained and tested on data reflecting the specific geographical, behavioural, contextual or functional setting within which they are intended to be used shall be presumed to comply with the relevant requirements laid down in Article 10(4).

+ 2. High-risk AI systems that have been certified or for which a statement of conformity has been issued under a cybersecurity scheme pursuant to Regulation (EU) 2019/881 and the references of which have been published in the Official Journal of the European Union shall be presumed to comply with the cybersecurity requirements set out in Article 15 of this Regulation in so far as the cybersecurity certificate or statement of conformity or parts thereof cover those requirements.

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 77/144

+ Article 43 Conformity assessment

+ 1. For high-risk AI systems listed in point 1 of Annex III, where, in demonstrating the compliance of a high-risk AI system with the requirements set out in Section 2, the provider has applied harmonised standards referred to in Article 40, or, where applicable, common specifications referred to in Article 41, the provider shall opt for one of the following conformity assessment procedures based on:

+ (a) the internal control referred to in Annex VI; or

+ (b) the assessment of the quality management system and the assessment of the technical documentation, with the involvement of a notified body, referred to in Annex VII.

+ In demonstrating the compliance of a high-risk AI system with the requirements set out in Section 2, the provider shall follow the conformity assessment procedure set out in Annex VII where:

+ (a) harmonised standards referred to in Article 40 do not exist, and common specifications referred to in Article 41 are not available;

+ (b) the provider has not applied, or has applied only part of, the harmonised standard;

+ (c) the common specifications referred to in point (a) exist, but the provider has not applied them;

+ (d) one or more of the harmonised standards referred to in point (a) has been published with a restriction, and only on the part of the standard that was restricted.

+ For the purposes of the conformity assessment procedure referred to in Annex VII, the provider may choose any of the notified bodies. However, where the high-risk AI system is intended to be put into service by law enforcement, immigration or asylum authorities or by Union institutions, bodies, offices or agencies, the market surveillance authority referred to in Article 74(8) or (9), as applicable, shall act as a notified body.

+ 2. For high-risk AI systems referred to in points 2 to 8 of Annex III, providers shall follow the conformity assessment procedure based on internal control as referred to in Annex VI, which does not provide for the involvement of a notified body.

+ 3. For high-risk AI systems covered by the Union harmonisation legislation listed in Section A of Annex I, the provider shall follow the relevant conformity assessment procedure as required under those legal acts. The requirements set out in Section 2 of this Chapter shall apply to those high-risk AI systems and shall be part of that assessment. Points 4.3., 4.4., 4.5. and the fifth paragraph of point 4.6 of Annex VII shall also apply.

+ For the purposes of that assessment, notified bodies which have been notified under those legal acts shall be entitled to control the conformity of the high-risk AI systems with the requirements set out in Section 2, provided that the compliance of those notified bodies with requirements laid down in Article 31(4), (5), (10) and (11) has been assessed in the context of the notification procedure under those legal acts.

+ Where a legal act listed in Section A of Annex I enables the product manufacturer to opt out from a third-party conformity assessment, provided that that manufacturer has applied all harmonised standards covering all the relevant requirements, that manufacturer may use that option only if it has also applied harmonised standards or, where applicable, common specifications referred to in Article 41, covering all requirements set out in Section 2 of this Chapter.

+ 4. High-risk AI systems that have already been subject to a conformity assessment procedure shall undergo a new conformity assessment procedure in the event of a substantial modification, regardless of whether the modified system is intended to be further distributed or continues to be used by the current deployer.

+ For high-risk AI systems that continue to learn after being placed on the market or put into service, changes to the high-risk AI system and its performance that have been pre-determined by the provider at the moment of the initial conformity assessment and are part of the information contained in the technical documentation referred to in point 2(f) of Annex IV, shall not constitute a substantial modification.

+ 5. The Commission is empowered to adopt delegated acts in accordance with Article 97 in order to amend Annexes VI and VII by updating them in light of technical progress.

+ 78/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 6. The Commission is empowered to adopt delegated acts in accordance with Article 97 in order to amend paragraphs 1 and 2 of this Article in order to subject high-risk AI systems referred to in points 2 to 8 of Annex III to the conformity assessment procedure referred to in Annex VII or parts thereof. The Commission shall adopt such delegated acts taking into account the effectiveness of the conformity assessment procedure based on internal control referred to in Annex VI in preventing or minimising the risks to health and safety and protection of fundamental rights posed by such systems, as well as the availability of adequate capacities and resources among notified bodies.

+ Article 44 Certificates

+ 1. Certificates issued by notified bodies in accordance with Annex VII shall be drawn-up in a language which can be easily understood by the relevant authorities in the Member State in which the notified body is established.

+ 2. Certificates shall be valid for the period they indicate, which shall not exceed five years for AI systems covered by Annex I, and four years for AI systems covered by Annex III. At the request of the provider, the validity of a certificate may be extended for further periods, each not exceeding five years for AI systems covered by Annex I, and four years for AI systems covered by Annex III, based on a re-assessment in accordance with the applicable conformity assessment procedures. Any supplement to a certificate shall remain valid, provided that the certificate which it supplements is valid.

+ 3. Where a notified body finds that an AI system no longer meets the requirements set out in Section 2, it shall, taking account of the principle of proportionality, suspend or withdraw the certificate issued or impose restrictions on it, unless compliance with those requirements is ensured by appropriate corrective action taken by the provider of the system within an appropriate deadline set by the notified body. The notified body shall give reasons for its decision.

+ An appeal procedure against decisions of the notified bodies, including on conformity certificates issued, shall be available.

+ Article 45 Information obligations of notified bodies

+ 1. Notified bodies shall inform the notifying authority of the following:

+ (a) any Union technical documentation assessment certificates, any supplements to those certificates, and any quality management system approvals issued in accordance with the requirements of Annex VII;

+ (b) any refusal, restriction, suspension or withdrawal of a Union technical documentation assessment certificate or a quality management system approval issued in accordance with the requirements of Annex VII;

+ (c) any circumstances affecting the scope of or conditions for notification;

+ (d) any request for information which they have received from market surveillance authorities regarding conformity assessment activities;

+ (e) on request, conformity assessment activities performed within the scope of their notification and any other activity performed, including cross-border activities and subcontracting.

+ 2. Each notified body shall inform the other notified bodies of:

+ (a) quality management system approvals which it has refused, suspended or withdrawn, and, upon request, of quality system approvals which it has issued;

+ (b) Union technical documentation assessment certificates or any supplements thereto which it has refused, withdrawn, suspended or otherwise restricted, and, upon request, of the certificates and/or supplements thereto which it has issued.

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 79/144

+ 3. Each notified body shall provide the other notified bodies carrying out similar conformity assessment activities covering the same types of AI systems with relevant information on issues relating to negative and, on request, positive conformity assessment results.

+ 4. Notified bodies shall safeguard the confidentiality of the information that they obtain, in accordance with Article 78.

+ Article 46

+ Derogation from conformity assessment procedure

+ 1. By way of derogation from Article 43 and upon a duly justified request, any market surveillance authority may authorise the placing on the market or the putting into service of specific high-risk AI systems within the territory of the Member State concerned, for exceptional reasons of public security or the protection of life and health of persons, environmental protection or the protection of key industrial and infrastructural assets. That authorisation shall be for a limited period while the necessary conformity assessment procedures are being carried out, taking into account the exceptional reasons justifying the derogation. The completion of those procedures shall be undertaken without undue delay.

+ 2. In a duly justified situation of urgency for exceptional reasons of public security or in the case of specific, substantial and imminent threat to the life or physical safety of natural persons, law-enforcement authorities or civil protection authorities may put a specific high-risk AI system into service without the authorisation referred to in paragraph 1, provided that such authorisation is requested during or after the use without undue delay. If the authorisation referred to in paragraph 1 is refused, the use of the high-risk AI system shall be stopped with immediate effect and all the results and outputs of such use shall be immediately discarded.

+ 3. The authorisation referred to in paragraph 1 shall be issued only if the market surveillance authority concludes that the high-risk AI system complies with the requirements of Section 2. The market surveillance authority shall inform the Commission and the other Member States of any authorisation issued pursuant to paragraphs 1 and 2. This obligation shall not cover sensitive operational data in relation to the activities of law-enforcement authorities.

+ 4. Where, within 15 calendar days of receipt of the information referred to in paragraph 3, no objection has been raised by either a Member State or the Commission in respect of an authorisation issued by a market surveillance authority of a Member State in accordance with paragraph 1, that authorisation shall be deemed justified.

+ 5. Where, within 15 calendar days of receipt of the notification referred to in paragraph 3, objections are raised by a Member State against an authorisation issued by a market surveillance authority of another Member State, or where the Commission considers the authorisation to be contrary to Union law, or the conclusion of the Member States regarding the compliance of the system as referred to in paragraph 3 to be unfounded, the Commission shall, without delay, enter into consultations with the relevant Member State. The operators concerned shall be consulted and have the possibility to present their views. Having regard thereto, the Commission shall decide whether the authorisation is justified. The Commission shall address its decision to the Member State concerned and to the relevant operators.

+ 6. Where the Commission considers the authorisation unjustified, it shall be withdrawn by the market surveillance authority of the Member State concerned.

+ 7. For high-risk AI systems related to products covered by Union harmonisation legislation listed in Section A of Annex I, only the derogations from the conformity assessment established in that Union harmonisation legislation shall apply.

+ Article 47 EU declaration of conformity

+ 1. The provider shall draw up a written machine readable, physical or electronically signed EU declaration of conformity for each high-risk AI system, and keep it at the disposal of the national competent authorities for 10 years after the high-risk AI system has been placed on the market or put into service. The EU declaration of conformity shall identify the high-risk AI system for which it has been drawn up. A copy of the EU declaration of conformity shall be submitted to the relevant national competent authorities upon request.

+ 80/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 2. The EU declaration of conformity shall state that the high-risk AI system concerned meets the requirements set out in Section 2. The EU declaration of conformity shall contain the information set out in Annex V, and shall be translated into a language that can be easily understood by the national competent authorities of the Member States in which the high-risk AI system is placed on the market or made available.

+ 3. Where high-risk AI systems are subject to other Union harmonisation legislation which also requires an EU declaration of conformity, a single EU declaration of conformity shall be drawn up in respect of all Union law applicable to the high-risk AI system. The declaration shall contain all the information required to identify the Union harmonisation legislation to which the declaration relates.

+ 4. By drawing up the EU declaration of conformity, the provider shall assume responsibility for compliance with the requirements set out in Section 2. The provider shall keep the EU declaration of conformity up-to-date as appropriate.

+ 5. The Commission is empowered to adopt delegated acts in accordance with Article 97 in order to amend Annex V by updating the content of the EU declaration of conformity set out in that Annex, in order to introduce elements that become necessary in light of technical progress.

+ Article 48 CE marking

+ 1. The CE marking shall be subject to the general principles set out in Article 30 of Regulation (EC) No 765/2008.

+ 2. For high-risk AI systems provided digitally, a digital CE marking shall be used, only if it can easily be accessed via the interface from which that system is accessed or via an easily accessible machine-readable code or other electronic means.

+ 3. The CE marking shall be affixed visibly, legibly and indelibly for high-risk AI systems. Where that is not possible or not warranted on account of the nature of the high-risk AI system, it shall be affixed to the packaging or to the accompanying documentation, as appropriate.

+ 4. Where applicable, the CE marking shall be followed by the identification number of the notified body responsible for the conformity assessment procedures set out in Article 43. The identification number of the notified body shall be affixed by the body itself or, under its instructions, by the provider or by the provider's authorised representative. The identification number shall also be indicated in any promotional material which mentions that the high-risk AI system fulfils the requirements for CE marking.

+ 5. Where high-risk AI systems are subject to other Union law which also provides for the affixing of the CE marking, the CE marking shall indicate that the high-risk AI system also fulfil the requirements of that other law.

+ Article 49 Registration

+ 1. Before placing on the market or putting into service a high-risk AI system listed in Annex III, with the exception of high-risk AI systems referred to in point 2 of Annex III, the provider or, where applicable, the authorised representative shall register themselves and their system in the EU database referred to in Article 71.

+ 2. Before placing on the market or putting into service an AI system for which the provider has concluded that it is not high-risk according to Article 6(3), that provider or, where applicable, the authorised representative shall register themselves and that system in the EU database referred to in Article 71.

+ 3. Before putting into service or using a high-risk AI system listed in Annex III, with the exception of high-risk AI systems listed in point 2 of Annex III, deployers that are public authorities, Union institutions, bodies, offices or agencies or persons acting on their behalf shall register themselves, select the system and register its use in the EU database referred to in Article 71.

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 81/144

+ EN

+ OJ L, 12.7.2024

+ 4. For high-risk AI systems referred to in points 1, 6 and 7 of Annex III, in the areas of law enforcement, migration, asylum and border control management, the registration referred to in paragraphs 1, 2 and 3 of this Article shall be in a secure non-public section of the EU database referred to in Article 71 and shall include only the following information, as applicable, referred to in:

+ (a) Section A, points 1 to 10, of Annex VIII, with the exception of points 6, 8 and 9;

+ Only the Commission and national authorities referred to in Article 74(8) shall have access to the respective restricted sections of the EU database listed in the first subparagraph of this paragraph.

+ 5. High-risk AI systems referred to in point 2 of Annex III shall be registered at national level.

+ CHAPTER IV TRANSPARENCY OBLIGATIONS FOR PROVIDERS AND DEPLOYERS OF CERTAIN AI SYSTEMS

+ Article 50 Transparency obligations for providers and deployers of certain AI systems

+ 1. Providers shall ensure that AI systems intended to interact directly with natural persons are designed and developed in such a way that the natural persons concerned are informed that they are interacting with an AI system, unless this is obvious from the point of view of a natural person who is reasonably well-informed, observant and circumspect, taking into account the circumstances and the context of use. This obligation shall not apply to AI systems authorised by law to detect, prevent, investigate or prosecute criminal offences, subject to appropriate safeguards for the rights and freedoms of third parties, unless those systems are available for the public to report a criminal offence.

+ 2. Providers of AI systems, including general-purpose AI systems, generating synthetic audio, image, video or text content, shall ensure that the outputs of the AI system are marked in a machine-readable format and detectable as artificially generated or manipulated. Providers shall ensure their technical solutions are effective, interoperable, robust and reliable as far as this is technically feasible, taking into account the specificities and limitations of various types of content, the costs of implementation and the generally acknowledged state of the art, as may be reflected in relevant technical standards. This obligation shall not apply to the extent the AI systems perform an assistive function for standard editing or do not substantially alter the input data provided by the deployer or the semantics thereof, or where authorised by law to detect, prevent, investigate or prosecute criminal offences.

+ 3. Deployers of an emotion recognition system or a biometric categorisation system shall inform the natural persons exposed thereto of the operation of the system, and shall process the personal data in accordance with Regulations (EU) 2016/679 and (EU) 2018/1725 and Directive (EU) 2016/680, as applicable. This obligation shall not apply to AI systems used for biometric categorisation and emotion recognition, which are permitted by law to detect, prevent or investigate criminal offences, subject to appropriate safeguards for the rights and freedoms of third parties, and in accordance with Union law.

+ 4. Deployers of an AI system that generates or manipulates image, audio or video content constituting a deep fake, shall disclose that the content has been artificially generated or manipulated. This obligation shall not apply where the use is authorised by law to detect, prevent, investigate or prosecute criminal offence. Where the content forms part of an evidently artistic, creative, satirical, fictional or analogous work or programme, the transparency obligations set out in this paragraph are limited to disclosure of the existence of such generated or manipulated content in an appropriate manner that does not hamper the display or enjoyment of the work.

+ Deployers of an AI system that generates or manipulates text which is published with the purpose of informing the public on matters of public interest shall disclose that the text has been artificially generated or manipulated. This obligation shall not apply where the use is authorised by law to detect, prevent, investigate or prosecute criminal offences or where the AI-generated content has undergone a process of human review or editorial control and where a natural or legal person holds editorial responsibility for the publication of the content.

+ 82/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ OJ L, 12.7.2024

+ EN

+ 5. The information referred to in paragraphs 1 to 4 shall be provided to the natural persons concerned in a clear and distinguishable manner at the latest at the time of the first interaction or exposure. The information shall conform to the applicable accessibility requirements.

+ 6. Paragraphs 1 to 4 shall not affect the requirements and obligations set out in Chapter III, and shall be without prejudice to other transparency obligations laid down in Union or national law for deployers of AI systems.

+ 7. The AI Office shall encourage and facilitate the drawing up of codes of practice at Union level to facilitate the effective implementation of the obligations regarding the detection and labelling of artificially generated or manipulated content. The Commission may adopt implementing acts to approve those codes of practice in accordance with the procedure laid down in Article 56 (6). If it deems the code is not adequate, the Commission may adopt an implementing act specifying common rules for the implementation of those obligations in accordance with the examination procedure laid down in Article 98(2).

+ CHAPTER V GENERAL-PURPOSE AI MODELS

+ SECTION 1

+ Classification rules

+ Article 51 Classification of general-purpose AI models as general-purpose AI models with systemic risk

+ 1. A general-purpose AI model shall be classified as a general-purpose AI model with systemic risk if it meets any of the following conditions:

+ (a) it has high impact capabilities evaluated on the basis of appropriate technical tools and methodologies, including indicators and benchmarks;

+ (b) based on a decision of the Commission, ex officio or following a qualified alert from the scientific panel, it has capabilities or an impact equivalent to those set out in point (a) having regard to the criteria set out in Annex XIII.

+ 2. A general-purpose AI model shall be presumed to have high impact capabilities pursuant to paragraph 1, point (a), when the cumulative amount of computation used for its training measured in floating point operations is greater than 1025.

+ 3. The Commission shall adopt delegated acts in accordance with Article 97 to amend the thresholds listed in paragraphs 1 and 2 of this Article, as well as to supplement benchmarks and indicators in light of evolving technological developments, such as algorithmic improvements or increased hardware efficiency, when necessary, for these thresholds to reflect the state of the art.

+ Article 52 Procedure

+ 1. Where a general-purpose AI model meets the condition referred to in Article 51(1), point (a), the relevant provider shall notify the Commission without delay and in any event within two weeks after that requirement is met or it becomes known that it will be met. That notification shall include the information necessary to demonstrate that the relevant requirement has been met. If the Commission becomes aware of a general-purpose AI model presenting systemic risks of which it has not been notified, it may decide to designate it as a model with systemic risk.

+ 2. The provider of a general-purpose AI model that meets the condition referred to in Article 51(1), point (a), may present, with its notification, sufficiently substantiated arguments to demonstrate that, exceptionally, although it meets that requirement, the general-purpose AI model does not present, due to its specific characteristics, systemic risks and therefore should not be classified as a general-purpose AI model with systemic risk.

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 83/144

+ 3. Where the Commission concludes that the arguments submitted pursuant to paragraph 2 are not sufficiently substantiated and the relevant provider was not able to demonstrate that the general-purpose AI model does not present, due to its specific characteristics, systemic risks, it shall reject those arguments, and the general-purpose AI model shall be considered to be a general-purpose AI model with systemic risk.

+ 4. The Commission may designate a general-purpose AI model as presenting systemic risks, ex officio or following a qualified alert from the scientific panel pursuant to Article 90(1), point (a), on the basis of criteria set out in Annex XIII.

+ The Commission is empowered to adopt delegated acts in accordance with Article 97 in order to amend Annex XIII by specifying and updating the criteria set out in that Annex.

+ 5. Upon a reasoned request of a provider whose model has been designated as a general-purpose AI model with systemic risk pursuant to paragraph 4, the Commission shall take the request into account and may decide to reassess whether the general-purpose AI model can still be considered to present systemic risks on the basis of the criteria set out in Annex XIII. Such a request shall contain objective, detailed and new reasons that have arisen since the designation decision. Providers may request reassessment at the earliest six months after the designation decision. Where the Commission, following its reassessment, decides to maintain the designation as a general-purpose AI model with systemic risk, providers may request reassessment at the earliest six months after that decision.

+ 6. The Commission shall ensure that a list of general-purpose AI models with systemic risk is published and shall keep that list up to date, without prejudice to the need to observe and protect intellectual property rights and confidential business information or trade secrets in accordance with Union and national law.

+ SECTION 2

+ Obligations for providers of general-purpose AI models

+ Article 53 Obligations for providers of general-purpose AI models

+ 1. Providers of general-purpose AI models shall:

+ (a) draw up and keep up-to-date the technical documentation of the model, including its training and testing process and the results of its evaluation, which shall contain, at a minimum, the information set out in Annex XI for the purpose of providing it, upon request, to the AI Office and the national competent authorities;

+ (b) draw up, keep up-to-date and make available information and documentation to providers of AI systems who intend to integrate the general-purpose AI model into their AI systems. Without prejudice to the need to observe and protect intellectual property rights and confidential business information or trade secrets in accordance with Union and national law, the information and documentation shall:

+ (i) enable providers of AI systems to have a good understanding of the capabilities and limitations of the general-purpose AI model and to comply with their obligations pursuant to this Regulation; and

+ (ii) contain, at a minimum, the elements set out in Annex XII;

+ (c) put in place a policy to comply with Union law on copyright and related rights, and in particular to identify and comply with, including through state-of-the-art technologies, a reservation of rights expressed pursuant to Article 4(3) of Directive (EU) 2019/790;

+ (d) draw up and make publicly available a sufficiently detailed summary about the content used for training of the general-purpose AI model, according to a template provided by the AI Office.

+ 84/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 2. The obligations set out in paragraph 1, points (a) and (b), shall not apply to providers of AI models that are released under a free and open-source licence that allows for the access, usage, modification, and distribution of the model, and whose parameters, including the weights, the information on the model architecture, and the information on model usage, are made publicly available. This exception shall not apply to general-purpose AI models with systemic risks.

+ 3. Providers of general-purpose AI models shall cooperate as necessary with the Commission and the national competent authorities in the exercise of their competences and powers pursuant to this Regulation.

+ 4. Providers of general-purpose AI models may rely on codes of practice within the meaning of Article 56 to demonstrate compliance with the obligations set out in paragraph 1 of this Article, until a harmonised standard is published. Compliance with European harmonised standards grants providers the presumption of conformity to the extent that those standards cover those obligations. Providers of general-purpose AI models who do not adhere to an approved code of practice or do not comply with a European harmonised standard shall demonstrate alternative adequate means of compliance for assessment by the Commission.

+ 5. For the purpose of facilitating compliance with Annex XI, in particular points 2 (d) and (e) thereof, the Commission is empowered to adopt delegated acts in accordance with Article 97 to detail measurement and calculation methodologies with a view to allowing for comparable and verifiable documentation.

+ 6. The Commission is empowered to adopt delegated acts in accordance with Article 97(2) to amend Annexes XI and XII in light of evolving technological developments.

+ 7. Any information or documentation obtained pursuant to this Article, including trade secrets, shall be treated in accordance with the confidentiality obligations set out in Article 78.

+ Article 54 Authorised representatives of providers of general-purpose AI models

+ 1. Prior to placing a general-purpose AI model on the Union market, providers established in third countries shall, by written mandate, appoint an authorised representative which is established in the Union.

+ 2. The provider shall enable its authorised representative to perform the tasks specified in the mandate received from the provider.

+ 3. The authorised representative shall perform the tasks specified in the mandate received from the provider. It shall provide a copy of the mandate to the AI Office upon request, in one of the official languages of the institutions of the Union. For the purposes of this Regulation, the mandate shall empower the authorised representative to carry out the following tasks:

+ (a) verify that the technical documentation specified in Annex XI has been drawn up and all obligations referred to in Article 53 and, where applicable, Article 55 have been fulfilled by the provider;

+ (b) keep a copy of the technical documentation specified in Annex XI at the disposal of the AI Office and national competent authorities, for a period of 10 years after the general-purpose AI model has been placed on the market, and the contact details of the provider that appointed the authorised representative;

+ (c) provide the AI Office, upon a reasoned request, with all the information and documentation, including that referred to in point (b), necessary to demonstrate compliance with the obligations in this Chapter;

+ (d) cooperate with the AI Office and competent authorities, upon a reasoned request, in any action they take in relation to the general-purpose AI model, including when the model is integrated into AI systems placed on the market or put into service in the Union.

+ 4. The mandate shall empower the authorised representative to be addressed, in addition to or instead of the provider, by the AI Office or the competent authorities, on all issues related to ensuring compliance with this Regulation.

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 85/144

+ 5. The authorised representative shall terminate the mandate if it considers or has reason to consider the provider to be acting contrary to its obligations pursuant to this Regulation. In such a case, it shall also immediately inform the AI Office about the termination of the mandate and the reasons therefor.

+ 6. The obligation set out in this Article shall not apply to providers of general-purpose AI models that are released under a free and open-source licence that allows for the access, usage, modification, and distribution of the model, and whose parameters, including the weights, the information on the model architecture, and the information on model usage, are made publicly available, unless the general-purpose AI models present systemic risks.

+ SECTION 3

+ Obligations of providers of general-purpose AI models with systemic risk

+ Article 55 Obligations of providers of general-purpose AI models with systemic risk

+ 1. In addition to the obligations listed in Articles 53 and 54, providers of general-purpose AI models with systemic risk shall:

+ (a) perform model evaluation in accordance with standardised protocols and tools reflecting the state of the art, including conducting and documenting adversarial testing of the model with a view to identifying and mitigating systemic risks;

+ (b) assess and mitigate possible systemic risks at Union level, including their sources, that may stem from the development, the placing on the market, or the use of general-purpose AI models with systemic risk;

+ (c) keep track of, document, and report, without undue delay, to the AI Office and, as appropriate, to national competent authorities, relevant information about serious incidents and possible corrective measures to address them;

+ (d) ensure an adequate level of cybersecurity protection for the general-purpose AI model with systemic risk and the physical infrastructure of the model.

+ 2. Providers of general-purpose AI models with systemic risk may rely on codes of practice within the meaning of Article 56 to demonstrate compliance with the obligations set out in paragraph 1 of this Article, until a harmonised standard is published. Compliance with European harmonised standards grants providers the presumption of conformity to the extent that those standards cover those obligations. Providers of general-purpose AI models with systemic risks who do not adhere to an approved code of practice or do not comply with a European harmonised standard shall demonstrate alternative adequate means of compliance for assessment by the Commission.

+ 3. Any information or documentation obtained pursuant to this Article, including trade secrets, shall be treated in accordance with the confidentiality obligations set out in Article 78.

+ SECTION 4

+ Codes of practice

+ Article 56 Codes of practice

+ 1. The AI Office shall encourage and facilitate the drawing up of codes of practice at Union level in order to contribute to the proper application of this Regulation, taking into account international approaches.

+ 2. The AI Office and the Board shall aim to ensure that the codes of practice cover at least the obligations provided for in Articles 53 and 55, including the following issues:

+ 86/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ (a) the means to ensure that the information referred to in Article 53(1), points (a) and (b), is kept up to date in light of market and technological developments;

+ (b) the adequate level of detail for the summary about the content used for training;

+ (c) the identification of the type and nature of the systemic risks at Union level, including their sources, where appropriate;

+ (d) the measures, procedures and modalities for the assessment and management of the systemic risks at Union level, including the documentation thereof, which shall be proportionate to the risks, take into consideration their severity and probability and take into account the specific challenges of tackling those risks in light of the possible ways in which such risks may emerge and materialise along the AI value chain.

+ 3. The AI Office may invite all providers of general-purpose AI models, as well as relevant national competent authorities, to participate in the drawing-up of codes of practice. Civil society organisations, industry, academia and other relevant stakeholders, such as downstream providers and independent experts, may support the process.

+ 4. The AI Office and the Board shall aim to ensure that the codes of practice clearly set out their specific objectives and contain commitments or measures, including key performance indicators as appropriate, to ensure the achievement of those objectives, and that they take due account of the needs and interests of all interested parties, including affected persons, at Union level.

+ 5. The AI Office shall aim to ensure that participants to the codes of practice report regularly to the AI Office on the implementation of the commitments and the measures taken and their outcomes, including as measured against the key performance indicators as appropriate. Key performance indicators and reporting commitments shall reflect differences in size and capacity between various participants.

+ 6. The AI Office and the Board shall regularly monitor and evaluate the achievement of the objectives of the codes of practice by the participants and their contribution to the proper application of this Regulation. The AI Office and the Board shall assess whether the codes of practice cover the obligations provided for in Articles 53 and 55, and shall regularly monitor and evaluate the achievement of their objectives. They shall publish their assessment of the adequacy of the codes of practice.

+ The Commission may, by way of an implementing act, approve a code of practice and give it a general validity within the Union. That implementing act shall be adopted in accordance with the examination procedure referred to in Article 98(2).

+ 7. The AI Office may invite all providers of general-purpose AI models to adhere to the codes of practice. For providers of general-purpose AI models not presenting systemic risks this adherence may be limited to the obligations provided for in Article 53, unless they declare explicitly their interest to join the full code.

+ 8. The AI Office shall, as appropriate, also encourage and facilitate the review and adaptation of the codes of practice, in particular in light of emerging standards. The AI Office shall assist in the assessment of available standards.

+ 9. Codes of practice shall be ready at the latest by 2 May 2025. The AI Office shall take the necessary steps, including inviting providers pursuant to paragraph 7.

+ If, by 2 August 2025, a code of practice cannot be finalised, or if the AI Office deems it is not adequate following its assessment under paragraph 6 of this Article, the Commission may provide, by means of implementing acts, common rules for the implementation of the obligations provided for in Articles 53 and 55, including the issues set out in paragraph 2 of this Article. Those implementing acts shall be adopted in accordance with the examination procedure referred to in Article 98(2).

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 87/144

+ EN

+ CHAPTER VI MEASURES IN SUPPORT OF INNOVATION

+ OJ L, 12.7.2024

+ Article 57 AI regulatory sandboxes

+ 1. Member States shall ensure that their competent authorities establish at least one AI regulatory sandbox at national level, which shall be operational by 2 August 2026. That sandbox may also be established jointly with the competent authorities of other Member States. The Commission may provide technical support, advice and tools for the establishment and operation of AI regulatory sandboxes.

+ The obligation under the first subparagraph may also be fulfilled by participating in an existing sandbox in so far as that participation provides an equivalent level of national coverage for the participating Member States.

+ 2. Additional AI regulatory sandboxes at regional or local level, or established jointly with the competent authorities of other Member States may also be established.

+ 3. The European Data Protection Supervisor may also establish an AI regulatory sandbox for Union institutions, bodies, offices and agencies, and may exercise the roles and the tasks of national competent authorities in accordance with this Chapter.

+ 4. Member States shall ensure that the competent authorities referred to in paragraphs 1 and 2 allocate sufficient resources to comply with this Article effectively and in a timely manner. Where appropriate, national competent authorities shall cooperate with other relevant authorities, and may allow for the involvement of other actors within the AI ecosystem. This Article shall not affect other regulatory sandboxes established under Union or national law. Member States shall ensure an appropriate level of cooperation between the authorities supervising those other sandboxes and the national competent authorities.

+ 5. AI regulatory sandboxes established under paragraph 1 shall provide for a controlled environment that fosters innovation and facilitates the development, training, testing and validation of innovative AI systems for a limited time before their being placed on the market or put into service pursuant to a specific sandbox plan agreed between the providers or prospective providers and the competent authority. Such sandboxes may include testing in real world conditions supervised therein.

+ 6. Competent authorities shall provide, as appropriate, guidance, supervision and support within the AI regulatory sandbox with a view to identifying risks, in particular to fundamental rights, health and safety, testing, mitigation measures, and their effectiveness in relation to the obligations and requirements of this Regulation and, where relevant, other Union and national law supervised within the sandbox.

+ 7. Competent authorities shall provide providers and prospective providers participating in the AI regulatory sandbox with guidance on regulatory expectations and how to fulfil the requirements and obligations set out in this Regulation.

+ Upon request of the provider or prospective provider of the AI system, the competent authority shall provide a written proof of the activities successfully carried out in the sandbox. The competent authority shall also provide an exit report detailing the activities carried out in the sandbox and the related results and learning outcomes. Providers may use such documentation to demonstrate their compliance with this Regulation through the conformity assessment process or relevant market surveillance activities. In this regard, the exit reports and the written proof provided by the national competent authority shall be taken positively into account by market surveillance authorities and notified bodies, with a view to accelerating conformity assessment procedures to a reasonable extent.

+ 8. Subject to the confidentiality provisions in Article 78, and with the agreement of the provider or prospective provider, the Commission and the Board shall be authorised to access the exit reports and shall take them into account, as appropriate, when exercising their tasks under this Regulation. If both the provider or prospective provider and the national competent authority explicitly agree, the exit report may be made publicly available through the single information platform referred to in this Article.

+ 9. The establishment of AI regulatory sandboxes shall aim to contribute to the following objectives:

+ (a) improving legal certainty to achieve regulatory compliance with this Regulation or, where relevant, other applicable Union and national law;

+ 88/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ OJ L, 12.7.2024

+ EN

+ (b) supporting the sharing of best practices through cooperation with the authorities involved in the AI regulatory sandbox;

+ (c) fostering innovation and competitiveness and facilitating the development of an AI ecosystem;

+ (d) contributing to evidence-based regulatory learning;

+ (e) facilitating and accelerating access to the Union market for AI systems, in particular when provided by SMEs, including start-ups.

+ 10. National competent authorities shall ensure that, to the extent the innovative AI systems involve the processing of personal data or otherwise fall under the supervisory remit of other national authorities or competent authorities providing or supporting access to data, the national data protection authorities and those other national or competent authorities are associated with the operation of the AI regulatory sandbox and involved in the supervision of those aspects to the extent of their respective tasks and powers.

+ 11. The AI regulatory sandboxes shall not affect the supervisory or corrective powers of the competent authorities supervising the sandboxes, including at regional or local level. Any significant risks to health and safety and fundamental rights identified during the development and testing of such AI systems shall result in an adequate mitigation. National competent authorities shall have the power to temporarily or permanently suspend the testing process, or the participation in the sandbox if no effective mitigation is possible, and shall inform the AI Office of such decision. National competent authorities shall exercise their supervisory powers within the limits of the relevant law, using their discretionary powers when implementing legal provisions in respect of a specific AI regulatory sandbox project, with the objective of supporting innovation in AI in the Union.

+ 12. Providers and prospective providers participating in the AI regulatory sandbox shall remain liable under applicable Union and national liability law for any damage inflicted on third parties as a result of the experimentation taking place in the sandbox. However, provided that the prospective providers observe the specific plan and the terms and conditions for their participation and follow in good faith the guidance given by the national competent authority, no administrative fines shall be imposed by the authorities for infringements of this Regulation. Where other competent authorities responsible for other Union and national law were actively involved in the supervision of the AI system in the sandbox and provided guidance for compliance, no administrative fines shall be imposed regarding that law.

+ 13. The AI regulatory sandboxes shall be designed and implemented in such a way that, where relevant, they facilitate cross-border cooperation between national competent authorities.

+ 14. National competent authorities shall coordinate their activities and cooperate within the framework of the Board.

+ 15. National competent authorities shall inform the AI Office and the Board of the establishment of a sandbox, and may ask them for support and guidance. The AI Office shall make publicly available a list of planned and existing sandboxes and keep it up to date in order to encourage more interaction in the AI regulatory sandboxes and cross-border cooperation.

+ 16. National competent authorities shall submit annual reports to the AI Office and to the Board, from one year after the establishment of the AI regulatory sandbox and every year thereafter until its termination, and a final report. Those reports shall provide information on the progress and results of the implementation of those sandboxes, including best practices, incidents, lessons learnt and recommendations on their setup and, where relevant, on the application and possible revision of this Regulation, including its delegated and implementing acts, and on the application of other Union law supervised by the competent authorities within the sandbox. The national competent authorities shall make those annual reports or abstracts thereof available to the public, online. The Commission shall, where appropriate, take the annual reports into account when exercising its tasks under this Regulation.

+ 17. The Commission shall develop a single and dedicated interface containing all relevant information related to AI regulatory sandboxes to allow stakeholders to interact with AI regulatory sandboxes and to raise enquiries with competent authorities, and to seek non-binding guidance on the conformity of innovative products, services, business models embedding AI technologies, in accordance with Article 62(1), point (c). The Commission shall proactively coordinate with national competent authorities, where relevant.

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 89/144

+ Article 58 Detailed arrangements for, and functioning of, AI regulatory sandboxes

+ 1. In order to avoid fragmentation across the Union, the Commission shall adopt implementing acts specifying the detailed arrangements for the establishment, development, implementation, operation and supervision of the AI regulatory sandboxes. The implementing acts shall include common principles on the following issues:

+ (a) eligibility and selection criteria for participation in the AI regulatory sandbox;

+ (b) procedures for the application, participation, monitoring, exiting from and termination of the AI regulatory sandbox, including the sandbox plan and the exit report;

+ (c) the terms and conditions applicable to the participants.

+ Those implementing acts shall be adopted in accordance with the examination procedure referred to in Article 98(2).

+ 2. The implementing acts referred to in paragraph 1 shall ensure:

+ (a) that AI regulatory sandboxes are open to any applying provider or prospective provider of an AI system who fulfils eligibility and selection criteria, which shall be transparent and fair, and that national competent authorities inform applicants of their decision within three months of the application;

+ (b) that AI regulatory sandboxes allow broad and equal access and keep up with demand for participation; providers and prospective providers may also submit applications in partnerships with deployers and other relevant third parties;

+ (c) that the detailed arrangements for, and conditions concerning AI regulatory sandboxes support, to the best extent possible, flexibility for national competent authorities to establish and operate their AI regulatory sandboxes;

+ (d) that access to the AI regulatory sandboxes is free of charge for SMEs, including start-ups, without prejudice to exceptional costs that national competent authorities may recover in a fair and proportionate manner;

+ (e) that they facilitate providers and prospective providers, by means of the learning outcomes of the AI regulatory sandboxes, in complying with conformity assessment obligations under this Regulation and the voluntary application of the codes of conduct referred to in Article 95;

+ (f) that AI regulatory sandboxes facilitate the involvement of other relevant actors within the AI ecosystem, such as notified bodies and standardisation organisations, SMEs, including start-ups, enterprises, innovators, testing and experimenta tion facilities, research and experimentation labs and European Digital Innovation Hubs, centres of excellence, individual researchers, in order to allow and facilitate cooperation with the public and private sectors;

+ (g) that procedures, processes and administrative requirements for application, selection, participation and exiting the AI regulatory sandbox are simple, easily intelligible, and clearly communicated in order to facilitate the participation of SMEs, including start-ups, with limited legal and administrative capacities and are streamlined across the Union, in order to avoid fragmentation and that participation in an AI regulatory sandbox established by a Member State, or by the European Data Protection Supervisor is mutually and uniformly recognised and carries the same legal effects across the Union;

+ (h) that participation in the AI regulatory sandbox is limited to a period that is appropriate to the complexity and scale of the project and that may be extended by the national competent authority;

+ (i) that AI regulatory sandboxes facilitate the development of tools and infrastructure for testing, benchmarking, assessing and explaining dimensions of AI systems relevant for regulatory learning, such as accuracy, robustness and cybersecurity, as well as measures to mitigate risks to fundamental rights and society at large.

+ 3. Prospective providers in the AI regulatory sandboxes, in particular SMEs and start-ups, shall be directed, where relevant, to pre-deployment services such as guidance on the implementation of this Regulation, to other value-adding services such as help with standardisation documents and certification, testing and experimentation facilities, European Digital Innovation Hubs and centres of excellence.

+ 90/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 4. Where national competent authorities consider authorising testing in real world conditions supervised within the framework of an AI regulatory sandbox to be established under this Article, they shall specifically agree the terms and conditions of such testing and, in particular, the appropriate safeguards with the participants, with a view to protecting fundamental rights, health and safety. Where appropriate, they shall cooperate with other national competent authorities with a view to ensuring consistent practices across the Union.

+ Article 59

+ Further processing of personal data for developing certain AI systems in the public interest in the AI regulatory sandbox

+ 1. In the AI regulatory sandbox, personal data lawfully collected for other purposes may be processed solely for the purpose of developing, training and testing certain AI systems in the sandbox when all of the following conditions are met:

+ (a) AI systems shall be developed for safeguarding substantial public interest by a public authority or another natural or legal person and in one or more of the following areas:

+ (i) public safety and public health, including disease detection, diagnosis prevention, control and treatment and improvement of health care systems;

+ (ii) a high level of protection and improvement of the quality of the environment, protection of biodiversity, protection against pollution, green transition measures, climate change mitigation and adaptation measures;

+ (iii) energy sustainability;

+ (iv) safety and resilience of transport systems and mobility, critical infrastructure and networks;

+ (v) efficiency and quality of public administration and public services;

+ (b) the data processed are necessary for complying with one or more of the requirements referred to in Chapter III, Section 2 where those requirements cannot effectively be fulfilled by processing anonymised, synthetic or other non-personal data;

+ (c) there are effective monitoring mechanisms to identify if any high risks to the rights and freedoms of the data subjects, as referred to in Article 35 of Regulation (EU) 2016/679 and in Article 39 of Regulation (EU) 2018/1725, may arise during the sandbox experimentation, as well as response mechanisms to promptly mitigate those risks and, where necessary, stop the processing;

+ (d) any personal data to be processed in the context of the sandbox are in a functionally separate, isolated and protected data processing environment under the control of the prospective provider and only authorised persons have access to those data;

+ (e) providers can further share the originally collected data only in accordance with Union data protection law; any personal data created in the sandbox cannot be shared outside the sandbox;

+ (f) any processing of personal data in the context of the sandbox neither leads to measures or decisions affecting the data subjects nor does it affect the application of their rights laid down in Union law on the protection of personal data;

+ (g) any personal data processed in the context of the sandbox are protected by means of appropriate technical and organisational measures and deleted once the participation in the sandbox has terminated or the personal data has reached the end of its retention period;

+ (h) the logs of the processing of personal data in the context of the sandbox are kept for the duration of the participation in the sandbox, unless provided otherwise by Union or national law;

+ (i) a complete and detailed description of the process and rationale behind the training, testing and validation of the AI system is kept together with the testing results as part of the technical documentation referred to in Annex IV;

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 91/144

+ (j) a short summary of the AI project developed in the sandbox, its objectives and expected results is published on the website of the competent authorities; this obligation shall not cover sensitive operational data in relation to the activities of law enforcement, border control, immigration or asylum authorities.

+ 2. For the purposes of the prevention, investigation, detection or prosecution of criminal offences or the execution of criminal penalties, including safeguarding against and preventing threats to public security, under the control and responsibility of law enforcement authorities, the processing of personal data in AI regulatory sandboxes shall be based on a specific Union or national law and subject to the same cumulative conditions as referred to in paragraph 1.

+ 3. Paragraph 1 is without prejudice to Union or national law which excludes processing of personal data for other purposes than those explicitly mentioned in that law, as well as to Union or national law laying down the basis for the processing of personal data which is necessary for the purpose of developing, testing or training of innovative AI systems or any other legal basis, in compliance with Union law on the protection of personal data.

+ Article 60 Testing of high-risk AI systems in real world conditions outside AI regulatory sandboxes

+ 1. Testing of high-risk AI systems in real world conditions outside AI regulatory sandboxes may be conducted by providers or prospective providers of high-risk AI systems listed in Annex III, in accordance with this Article and the real-world testing plan referred to in this Article, without prejudice to the prohibitions under Article 5.

+ The Commission shall, by means of implementing acts, specify the detailed elements of the real-world testing plan. Those implementing acts shall be adopted in accordance with the examination procedure referred to in Article 98(2).

+ This paragraph shall be without prejudice to Union or national law on the testing in real world conditions of high-risk AI systems related to products covered by Union harmonisation legislation listed in Annex I.

+ 2. Providers or prospective providers may conduct testing of high-risk AI systems referred to in Annex III in real world conditions at any time before the placing on the market or the putting into service of the AI system on their own or in partnership with one or more deployers or prospective deployers.

+ 3. The testing of high-risk AI systems in real world conditions under this Article shall be without prejudice to any ethical review that is required by Union or national law.

+ 4. Providers or prospective providers may conduct the testing in real world conditions only where all of the following conditions are met:

+ (a) the provider or prospective provider has drawn up a real-world testing plan and submitted it to the market surveillance authority in the Member State where the testing in real world conditions is to be conducted;

+ (b) the market surveillance authority in the Member State where the testing in real world conditions is to be conducted has approved the testing in real world conditions and the real-world testing plan; where the market surveillance authority has not provided an answer within 30 days, the testing in real world conditions and the real-world testing plan shall be understood to have been approved; where national law does not provide for a tacit approval, the testing in real world conditions shall remain subject to an authorisation;

+ (c) the provider or prospective provider, with the exception of providers or prospective providers of high-risk AI systems referred to in points 1, 6 and 7 of Annex III in the areas of law enforcement, migration, asylum and border control management, and high-risk AI systems referred to in point 2 of Annex III has registered the testing in real world conditions in accordance with Article 71(4) with a Union-wide unique single identification number and with the information specified in Annex IX; the provider or prospective provider of high-risk AI systems referred to in points 1, 6 and 7 of Annex III in the areas of law enforcement, migration, asylum and border control management, has registered the testing in real-world conditions in the secure non-public section of the EU database according to Article 49(4), point (d), with a Union-wide unique single identification number and with the information specified therein; the provider or prospective provider of high-risk AI systems referred to in point 2 of Annex III has registered the testing in real-world conditions in accordance with Article 49(5);

+ 92/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ (d) the provider or prospective provider conducting the testing in real world conditions is established in the Union or has appointed a legal representative who is established in the Union;

+ (e) data collected and processed for the purpose of the testing in real world conditions shall be transferred to third countries only provided that appropriate and applicable safeguards under Union law are implemented;

+ (f) the testing in real world conditions does not last longer than necessary to achieve its objectives and in any case not longer than six months, which may be extended for an additional period of six months, subject to prior notification by the provider or prospective provider to the market surveillance authority, accompanied by an explanation of the need for such an extension;

+ (g) the subjects of the testing in real world conditions who are persons belonging to vulnerable groups due to their age or disability, are appropriately protected;

+ (h) where a provider or prospective provider organises the testing in real world conditions in cooperation with one or more deployers or prospective deployers, the latter have been informed of all aspects of the testing that are relevant to their decision to participate, and given the relevant instructions for use of the AI system referred to in Article 13; the provider or prospective provider and the deployer or prospective deployer shall conclude an agreement specifying their roles and responsibilities with a view to ensuring compliance with the provisions for testing in real world conditions under this Regulation and under other applicable Union and national law;

+ (i) the subjects of the testing in real world conditions have given informed consent in accordance with Article 61, or in the case of law enforcement, where the seeking of informed consent would prevent the AI system from being tested, the testing itself and the outcome of the testing in the real world conditions shall not have any negative effect on the subjects, and their personal data shall be deleted after the test is performed;

+ (j) the testing in real world conditions is effectively overseen by the provider or prospective provider, as well as by deployers or prospective deployers through persons who are suitably qualified in the relevant field and have the necessary capacity, training and authority to perform their tasks;

+ (k) the predictions, recommendations or decisions of the AI system can be effectively reversed and disregarded.

+ 5. Any subjects of the testing in real world conditions, or their legally designated representative, as appropriate, may, without any resulting detriment and without having to provide any justification, withdraw from the testing at any time by revoking their informed consent and may request the immediate and permanent deletion of their personal data. The withdrawal of the informed consent shall not affect the activities already carried out.

+ 6. In accordance with Article 75, Member States shall confer on their market surveillance authorities the powers of requiring providers and prospective providers to provide information, of carrying out unannounced remote or on-site inspections, and of performing checks on the conduct of the testing in real world conditions and the related high-risk AI systems. Market surveillance authorities shall use those powers to ensure the safe development of testing in real world conditions.

+ 7. Any serious incident identified in the course of the testing in real world conditions shall be reported to the national market surveillance authority in accordance with Article 73. The provider or prospective provider shall adopt immediate mitigation measures or, failing that, shall suspend the testing in real world conditions until such mitigation takes place, or otherwise terminate it. The provider or prospective provider shall establish a procedure for the prompt recall of the AI system upon such termination of the testing in real world conditions.

+ 8. Providers or prospective providers shall notify the national market surveillance authority in the Member State where the testing in real world conditions is to be conducted of the suspension or termination of the testing in real world conditions and of the final outcomes.

+ 9. The provider or prospective provider shall be liable under applicable Union and national liability law for any damage caused in the course of their testing in real world conditions.

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 93/144

+ Article 61 Informed consent to participate in testing in real world conditions outside AI regulatory sandboxes

+ 1. For the purpose of testing in real world conditions under Article 60, freely-given informed consent shall be obtained from the subjects of testing prior to their participation in such testing and after their having been duly informed with concise, clear, relevant, and understandable information regarding:

+ (a) the nature and objectives of the testing in real world conditions and the possible inconvenience that may be linked to their participation;

+ (b) the conditions under which the testing in real world conditions is to be conducted, including the expected duration of the subject or subjects' participation;

+ (c) their rights, and the guarantees regarding their participation, in particular their right to refuse to participate in, and the right to withdraw from, testing in real world conditions at any time without any resulting detriment and without having to provide any justification;

+ (d) the arrangements for requesting the reversal or the disregarding of the predictions, recommendations or decisions of the AI system;

+ (e) the Union-wide unique single identification number of the testing in real world conditions in accordance with Article 60(4) point (c), and the contact details of the provider or its legal representative from whom further information can be obtained.

+ 2. The informed consent shall be dated and documented and a copy shall be given to the subjects of testing or their legal representative.

+ Article 62 Measures for providers and deployers, in particular SMEs, including start-ups

+ 1. Member States shall undertake the following actions:

+ (a) provide SMEs, including start-ups, having a registered office or a branch in the Union, with priority access to the AI regulatory sandboxes, to the extent that they fulfil the eligibility conditions and selection criteria; the priority access shall not preclude other SMEs, including start-ups, other than those referred to in this paragraph from access to the AI regulatory sandbox, provided that they also fulfil the eligibility conditions and selection criteria;

+ (b) organise specific awareness raising and training activities on the application of this Regulation tailored to the needs of SMEs including start-ups, deployers and, as appropriate, local public authorities;

+ (c) utilise existing dedicated channels and where appropriate, establish new ones for communication with SMEs including start-ups, deployers, other innovators and, as appropriate, local public authorities to provide advice and respond to queries about the implementation of this Regulation, including as regards participation in AI regulatory sandboxes;

+ (d) facilitate the participation of SMEs and other relevant stakeholders in the standardisation development process.

+ 2. The specific interests and needs of the SME providers, including start-ups, shall be taken into account when setting the fees for conformity assessment under Article 43, reducing those fees proportionately to their size, market size and other relevant indicators.

+ 3. The AI Office shall undertake the following actions:

+ (a) provide standardised templates for areas covered by this Regulation, as specified by the Board in its request;

+ (b) develop and maintain a single information platform providing easy to use information in relation to this Regulation for all operators across the Union;

+ 94/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ (c) organise appropriate communication campaigns to raise awareness about the obligations arising from this Regulation;

+ (d) evaluate and promote the convergence of best practices in public procurement procedures in relation to AI systems.

+ Article 63

+ Derogations for specific operators

+ 1. Microenterprises within the meaning of Recommendation 2003/361/EC may comply with certain elements of the quality management system required by Article 17 of this Regulation in a simplified manner, provided that they do not have partner enterprises or linked enterprises within the meaning of that Recommendation. For that purpose, the Commission shall develop guidelines on the elements of the quality management system which may be complied with in a simplified manner considering the needs of microenterprises, without affecting the level of protection or the need for compliance with the requirements in respect of high-risk AI systems.

+ 2. Paragraph 1 of this Article shall not be interpreted as exempting those operators from fulfilling any other requirements or obligations laid down in this Regulation, including those established in Articles 9, 10, 11, 12, 13, 14, 15, 72 and 73.

+ CHAPTER VII GOVERNANCE

+ SECTION 1

+ Governance at Union level

+ Article 64 AI Office 1. The Commission shall develop Union expertise and capabilities in the field of AI through the AI Office. 2. Member States shall facilitate the tasks entrusted to the AI Office, as reflected in this Regulation.

+ Article 65 Establishment and structure of the European Artificial Intelligence Board

+ 1. A European Artificial Intelligence Board (the `Board') is hereby established.

+ 2. The Board shall be composed of one representative per Member State. The European Data Protection Supervisor shall participate as observer. The AI Office shall also attend the Board's meetings, without taking part in the votes. Other national and Union authorities, bodies or experts may be invited to the meetings by the Board on a case by case basis, where the issues discussed are of relevance for them.

+ 3. Each representative shall be designated by their Member State for a period of three years, renewable once.

+ 4. Member States shall ensure that their representatives on the Board:

+ (a) have the relevant competences and powers in their Member State so as to contribute actively to the achievement of the Board's tasks referred to in Article 66;

+ (b) are designated as a single contact point vis-�-vis the Board and, where appropriate, taking into account Member States' needs, as a single contact point for stakeholders;

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 95/144

+ (c) are empowered to facilitate consistency and coordination between national competent authorities in their Member State as regards the implementation of this Regulation, including through the collection of relevant data and information for the purpose of fulfilling their tasks on the Board.

+ 5. The designated representatives of the Member States shall adopt the Board's rules of procedure by a two-thirds majority. The rules of procedure shall, in particular, lay down procedures for the selection process, the duration of the mandate of, and specifications of the tasks of, the Chair, detailed arrangements for voting, and the organisation of the Board's activities and those of its sub-groups.

+ 6. The Board shall establish two standing sub-groups to provide a platform for cooperation and exchange among market surveillance authorities and notifying authorities about issues related to market surveillance and notified bodies respectively.

+ The standing sub-group for market surveillance should act as the administrative cooperation group (ADCO) for this Regulation within the meaning of Article 30 of Regulation (EU) 2019/1020.

+ The Board may establish other standing or temporary sub-groups as appropriate for the purpose of examining specific issues. Where appropriate, representatives of the advisory forum referred to in Article 67 may be invited to such sub-groups or to specific meetings of those subgroups as observers.

+ 7. The Board shall be organised and operated so as to safeguard the objectivity and impartiality of its activities.

+ 8. The Board shall be chaired by one of the representatives of the Member States. The AI Office shall provide the secretariat for the Board, convene the meetings upon request of the Chair, and prepare the agenda in accordance with the tasks of the Board pursuant to this Regulation and its rules of procedure.

+ Article 66 Tasks of the Board

+ The Board shall advise and assist the Commission and the Member States in order to facilitate the consistent and effective application of this Regulation. To that end, the Board may in particular:

+ (a) contribute to the coordination among national competent authorities responsible for the application of this Regulation and, in cooperation with and subject to the agreement of the market surveillance authorities concerned, support joint activities of market surveillance authorities referred to in Article 74(11);

+ (b) collect and share technical and regulatory expertise and best practices among Member States;

+ (c) provide advice on the implementation of this Regulation, in particular as regards the enforcement of rules on general-purpose AI models;

+ (d) contribute to the harmonisation of administrative practices in the Member States, including in relation to the derogation from the conformity assessment procedures referred to in Article 46, the functioning of AI regulatory sandboxes, and testing in real world conditions referred to in Articles 57, 59 and 60;

+ (e) at the request of the Commission or on its own initiative, issue recommendations and written opinions on any relevant matters related to the implementation of this Regulation and to its consistent and effective application, including:

+ (i) on the development and application of codes of conduct and codes of practice pursuant to this Regulation, as well as of the Commission's guidelines;

+ (ii) the evaluation and review of this Regulation pursuant to Article 112, including as regards the serious incident reports referred to in Article 73, and the functioning of the EU database referred to in Article 71, the preparation of the delegated or implementing acts, and as regards possible alignments of this Regulation with the Union harmonisation legislation listed in Annex I;

+ (iii) on technical specifications or existing standards regarding the requirements set out in Chapter III, Section 2;

+ 96/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ (iv) on the use of harmonised standards or common specifications referred to in Articles 40 and 41;

+ (v) trends, such as European global competitiveness in AI, the uptake of AI in the Union, and the development of digital skills;

+ (vi) trends on the evolving typology of AI value chains, in particular on the resulting implications in terms of accountability;

+ (vii) on the potential need for amendment to Annex III in accordance with Article 7, and on the potential need for possible revision of Article 5 pursuant to Article 112, taking into account relevant available evidence and the latest developments in technology;

+ (f) support the Commission in promoting AI literacy, public awareness and understanding of the benefits, risks, safeguards and rights and obligations in relation to the use of AI systems;

+ (g) facilitate the development of common criteria and a shared understanding among market operators and competent authorities of the relevant concepts provided for in this Regulation, including by contributing to the development of benchmarks;

+ (h) cooperate, as appropriate, with other Union institutions, bodies, offices and agencies, as well as relevant Union expert groups and networks, in particular in the fields of product safety, cybersecurity, competition, digital and media services, financial services, consumer protection, data and fundamental rights protection;

+ (i) contribute to effective cooperation with the competent authorities of third countries and with international organisations;

+ (j) assist national competent authorities and the Commission in developing the organisational and technical expertise required for the implementation of this Regulation, including by contributing to the assessment of training needs for staff of Member States involved in implementing this Regulation;

+ (k) assist the AI Office in supporting national competent authorities in the establishment and development of AI regulatory sandboxes, and facilitate cooperation and information-sharing among AI regulatory sandboxes;

+ (l) contribute to, and provide relevant advice on, the development of guidance documents;

+ (m) advise the Commission in relation to international matters on AI;

+ (n) provide opinions to the Commission on the qualified alerts regarding general-purpose AI models;

+ (o) receive opinions by the Member States on qualified alerts regarding general-purpose AI models, and on national experiences and practices on the monitoring and enforcement of AI systems, in particular systems integrating the general-purpose AI models.

+ Article 67 Advisory forum

+ 1. An advisory forum shall be established to provide technical expertise and advise the Board and the Commission, and to contribute to their tasks under this Regulation.

+ 2. The membership of the advisory forum shall represent a balanced selection of stakeholders, including industry, start-ups, SMEs, civil society and academia. The membership of the advisory forum shall be balanced with regard to commercial and non-commercial interests and, within the category of commercial interests, with regard to SMEs and other undertakings.

+ 3. The Commission shall appoint the members of the advisory forum, in accordance with the criteria set out in paragraph 2, from amongst stakeholders with recognised expertise in the field of AI.

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 97/144

+ 4. The term of office of the members of the advisory forum shall be two years, which may be extended by up to no more than four years.

+ 5. The Fundamental Rights Agency, ENISA, the European Committee for Standardization (CEN), the European Committee for Electrotechnical Standardization (CENELEC), and the European Telecommunications Standards Institute (ETSI) shall be permanent members of the advisory forum.

+ 6. The advisory forum shall draw up its rules of procedure. It shall elect two co-chairs from among its members, in accordance with criteria set out in paragraph 2. The term of office of the co-chairs shall be two years, renewable once.

+ 7. The advisory forum shall hold meetings at least twice a year. The advisory forum may invite experts and other stakeholders to its meetings.

+ 8. The advisory forum may prepare opinions, recommendations and written contributions at the request of the Board or the Commission.

+ 9. The advisory forum may establish standing or temporary sub-groups as appropriate for the purpose of examining specific questions related to the objectives of this Regulation.

+ 10. The advisory forum shall prepare an annual report on its activities. That report shall be made publicly available.

+ Article 68 Scientific panel of independent experts

+ 1. The Commission shall, by means of an implementing act, make provisions on the establishment of a scientific panel of independent experts (the `scientific panel') intended to support the enforcement activities under this Regulation. That implementing act shall be adopted in accordance with the examination procedure referred to in Article 98(2).

+ 2. The scientific panel shall consist of experts selected by the Commission on the basis of up-to-date scientific or technical expertise in the field of AI necessary for the tasks set out in paragraph 3, and shall be able to demonstrate meeting all of the following conditions:

+ (a) having particular expertise and competence and scientific or technical expertise in the field of AI;

+ (b) independence from any provider of AI systems or general-purpose AI models;

+ (c) an ability to carry out activities diligently, accurately and objectively.

+ The Commission, in consultation with the Board, shall determine the number of experts on the panel in accordance with the required needs and shall ensure fair gender and geographical representation.

+ 3. The scientific panel shall advise and support the AI Office, in particular with regard to the following tasks:

+ (a) supporting the implementation and enforcement of this Regulation as regards general-purpose AI models and systems, in particular by:

+ (i) alerting the AI Office of possible systemic risks at Union level of general-purpose AI models, in accordance with Article 90;

+ (ii) contributing to the development of tools and methodologies for evaluating capabilities of general-purpose AI models and systems, including through benchmarks;

+ (iii) providing advice on the classification of general-purpose AI models with systemic risk;

+ (iv) providing advice on the classification of various general-purpose AI models and systems;

+ 98/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ (v) contributing to the development of tools and templates;

+ (b) supporting the work of market surveillance authorities, at their request;

+ (c) supporting cross-border market surveillance activities as referred to in Article 74(11), without prejudice to the powers of market surveillance authorities;

+ (d) supporting the AI Office in carrying out its duties in the context of the Union safeguard procedure pursuant to Article 81.

+ 4. The experts on the scientific panel shall perform their tasks with impartiality and objectivity, and shall ensure the confidentiality of information and data obtained in carrying out their tasks and activities. They shall neither seek nor take instructions from anyone when exercising their tasks under paragraph 3. Each expert shall draw up a declaration of interests, which shall be made publicly available. The AI Office shall establish systems and procedures to actively manage and prevent potential conflicts of interest.

+ 5. The implementing act referred to in paragraph 1 shall include provisions on the conditions, procedures and detailed arrangements for the scientific panel and its members to issue alerts, and to request the assistance of the AI Office for the performance of the tasks of the scientific panel.

+ Article 69 Access to the pool of experts by the Member States

+ 1. Member States may call upon experts of the scientific panel to support their enforcement activities under this Regulation.

+ 2. The Member States may be required to pay fees for the advice and support provided by the experts. The structure and the level of fees as well as the scale and structure of recoverable costs shall be set out in the implementing act referred to in Article 68(1), taking into account the objectives of the adequate implementation of this Regulation, cost-effectiveness and the necessity of ensuring effective access to experts for all Member States.

+ 3. The Commission shall facilitate timely access to the experts by the Member States, as needed, and ensure that the combination of support activities carried out by Union AI testing support pursuant to Article 84 and experts pursuant to this Article is efficiently organised and provides the best possible added value.

+ SECTION 2

+ National competent authorities

+ Article 70 Designation of national competent authorities and single points of contact

+ 1. Each Member State shall establish or designate as national competent authorities at least one notifying authority and at least one market surveillance authority for the purposes of this Regulation. Those national competent authorities shall exercise their powers independently, impartially and without bias so as to safeguard the objectivity of their activities and tasks, and to ensure the application and implementation of this Regulation. The members of those authorities shall refrain from any action incompatible with their duties. Provided that those principles are observed, such activities and tasks may be performed by one or more designated authorities, in accordance with the organisational needs of the Member State.

+ 2. Member States shall communicate to the Commission the identity of the notifying authorities and the market surveillance authorities and the tasks of those authorities, as well as any subsequent changes thereto. Member States shall make publicly available information on how competent authorities and single points of contact can be contacted, through electronic communication means by 2 August 2025. Member States shall designate a market surveillance authority to act as the single point of contact for this Regulation, and shall notify the Commission of the identity of the single point of contact. The Commission shall make a list of the single points of contact publicly available.

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 99/144

+ 3. Member States shall ensure that their national competent authorities are provided with adequate technical, financial and human resources, and with infrastructure to fulfil their tasks effectively under this Regulation. In particular, the national competent authorities shall have a sufficient number of personnel permanently available whose competences and expertise shall include an in-depth understanding of AI technologies, data and data computing, personal data protection, cybersecurity, fundamental rights, health and safety risks and knowledge of existing standards and legal requirements. Member States shall assess and, if necessary, update competence and resource requirements referred to in this paragraph on an annual basis.

+ 4. National competent authorities shall take appropriate measures to ensure an adequate level of cybersecurity.

+ 5. When performing their tasks, the national competent authorities shall act in accordance with the confidentiality obligations set out in Article 78.

+ 6. By 2 August 2025, and once every two years thereafter, Member States shall report to the Commission on the status of the financial and human resources of the national competent authorities, with an assessment of their adequacy. The Commission shall transmit that information to the Board for discussion and possible recommendations.

+ 7. The Commission shall facilitate the exchange of experience between national competent authorities.

+ 8. National competent authorities may provide guidance and advice on the implementation of this Regulation, in particular to SMEs including start-ups, taking into account the guidance and advice of the Board and the Commission, as appropriate. Whenever national competent authorities intend to provide guidance and advice with regard to an AI system in areas covered by other Union law, the national competent authorities under that Union law shall be consulted, as appropriate.

+ 9. Where Union institutions, bodies, offices or agencies fall within the scope of this Regulation, the European Data Protection Supervisor shall act as the competent authority for their supervision.

+ CHAPTER VIII EU DATABASE FOR HIGH-RISK AI SYSTEMS

+ Article 71 EU database for high-risk AI systems listed in Annex III

+ 1. The Commission shall, in collaboration with the Member States, set up and maintain an EU database containing information referred to in paragraphs 2 and 3 of this Article concerning high-risk AI systems referred to in Article 6(2) which are registered in accordance with Articles 49 and 60 and AI systems that are not considered as high-risk pursuant to Article 6(3) and which are registered in accordance with Article 6(4) and Article 49. When setting the functional specifications of such database, the Commission shall consult the relevant experts, and when updating the functional specifications of such database, the Commission shall consult the Board.

+ 2. The data listed in Sections A and B of Annex VIII shall be entered into the EU database by the provider or, where applicable, by the authorised representative.

+ 3. The data listed in Section C of Annex VIII shall be entered into the EU database by the deployer who is, or who acts on behalf of, a public authority, agency or body, in accordance with Article 49(3) and (4).

+ 4. With the exception of the section referred to in Article 49(4) and Article 60(4), point (c), the information contained in the EU database registered in accordance with Article 49 shall be accessible and publicly available in a user-friendly manner. The information should be easily navigable and machine-readable. The information registered in accordance with Article 60 shall be accessible only to market surveillance authorities and the Commission, unless the prospective provider or provider has given consent for also making the information accessible the public.

+ 5. The EU database shall contain personal data only in so far as necessary for collecting and processing information in accordance with this Regulation. That information shall include the names and contact details of natural persons who are responsible for registering the system and have the legal authority to represent the provider or the deployer, as applicable.

+ 100/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 6. The Commission shall be the controller of the EU database. It shall make available to providers, prospective providers and deployers adequate technical and administrative support. The EU database shall comply with the applicable accessibility requirements.

+ CHAPTER IX POST-MARKET MONITORING, INFORMATION SHARING AND MARKET SURVEILLANCE

+ SECTION 1

+ Post-market monitoring

+ Article 72 Post-market monitoring by providers and post-market monitoring plan for high-risk AI systems

+ 1. Providers shall establish and document a post-market monitoring system in a manner that is proportionate to the nature of the AI technologies and the risks of the high-risk AI system.

+ 2. The post-market monitoring system shall actively and systematically collect, document and analyse relevant data which may be provided by deployers or which may be collected through other sources on the performance of high-risk AI systems throughout their lifetime, and which allow the provider to evaluate the continuous compliance of AI systems with the requirements set out in Chapter III, Section 2. Where relevant, post-market monitoring shall include an analysis of the interaction with other AI systems. This obligation shall not cover sensitive operational data of deployers which are law-enforcement authorities.

+ 3. The post-market monitoring system shall be based on a post-market monitoring plan. The post-market monitoring plan shall be part of the technical documentation referred to in Annex IV. The Commission shall adopt an implementing act laying down detailed provisions establishing a template for the post-market monitoring plan and the list of elements to be included in the plan by 2 February 2026. That implementing act shall be adopted in accordance with the examination procedure referred to in Article 98(2).

+ 4. For high-risk AI systems covered by the Union harmonisation legislation listed in Section A of Annex I, where a post-market monitoring system and plan are already established under that legislation, in order to ensure consistency, avoid duplications and minimise additional burdens, providers shall have a choice of integrating, as appropriate, the necessary elements described in paragraphs 1, 2 and 3 using the template referred in paragraph 3 into systems and plans already existing under that legislation, provided that it achieves an equivalent level of protection.

+ The first subparagraph of this paragraph shall also apply to high-risk AI systems referred to in point 5 of Annex III placed on the market or put into service by financial institutions that are subject to requirements under Union financial services law regarding their internal governance, arrangements or processes.

+ SECTION 2

+ Sharing of information on serious incidents

+ Article 73 Reporting of serious incidents

+ 1. Providers of high-risk AI systems placed on the Union market shall report any serious incident to the market surveillance authorities of the Member States where that incident occurred.

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 101/144

+ 2. The report referred to in paragraph 1 shall be made immediately after the provider has established a causal link between the AI system and the serious incident or the reasonable likelihood of such a link, and, in any event, not later than 15 days after the provider or, where applicable, the deployer, becomes aware of the serious incident.

+ The period for the reporting referred to in the first subparagraph shall take account of the severity of the serious incident.

+ 3. Notwithstanding paragraph 2 of this Article, in the event of a widespread infringement or a serious incident as defined in Article 3, point (49)(b), the report referred to in paragraph 1 of this Article shall be provided immediately, and not later than two days after the provider or, where applicable, the deployer becomes aware of that incident.

+ 4. Notwithstanding paragraph 2, in the event of the death of a person, the report shall be provided immediately after the provider or the deployer has established, or as soon as it suspects, a causal relationship between the high-risk AI system and the serious incident, but not later than 10 days after the date on which the provider or, where applicable, the deployer becomes aware of the serious incident.

+ 5. Where necessary to ensure timely reporting, the provider or, where applicable, the deployer, may submit an initial report that is incomplete, followed by a complete report.

+ 6. Following the reporting of a serious incident pursuant to paragraph 1, the provider shall, without delay, perform the necessary investigations in relation to the serious incident and the AI system concerned. This shall include a risk assessment of the incident, and corrective action.

+ The provider shall cooperate with the competent authorities, and where relevant with the notified body concerned, during the investigations referred to in the first subparagraph, and shall not perform any investigation which involves altering the AI system concerned in a way which may affect any subsequent evaluation of the causes of the incident, prior to informing the competent authorities of such action.

+ 7. Upon receiving a notification related to a serious incident referred to in Article 3, point (49)(c), the relevant market surveillance authority shall inform the national public authorities or bodies referred to in Article 77(1). The Commission shall develop dedicated guidance to facilitate compliance with the obligations set out in paragraph 1 of this Article. That guidance shall be issued by 2 August 2025, and shall be assessed regularly.

+ 8. The market surveillance authority shall take appropriate measures, as provided for in Article 19 of Regulation (EU) 2019/1020, within seven days from the date it received the notification referred to in paragraph 1 of this Article, and shall follow the notification procedures as provided in that Regulation.

+ 9. For high-risk AI systems referred to in Annex III that are placed on the market or put into service by providers that are subject to Union legislative instruments laying down reporting obligations equivalent to those set out in this Regulation, the notification of serious incidents shall be limited to those referred to in Article 3, point (49)(c).

+ 10. For high-risk AI systems which are safety components of devices, or are themselves devices, covered by Regulations (EU) 2017/745 and (EU) 2017/746, the notification of serious incidents shall be limited to those referred to in Article 3, point (49)(c) of this Regulation, and shall be made to the national competent authority chosen for that purpose by the Member States where the incident occurred.

+ 11. National competent authorities shall immediately notify the Commission of any serious incident, whether or not they have taken action on it, in accordance with Article 20 of Regulation (EU) 2019/1020.

+ SECTION 3

+ Enforcement

+ Article 74 Market surveillance and control of AI systems in the Union market

+ 1. Regulation (EU) 2019/1020 shall apply to AI systems covered by this Regulation. For the purposes of the effective enforcement of this Regulation:

+ 102/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ (a) any reference to an economic operator under Regulation (EU) 2019/1020 shall be understood as including all operators identified in Article 2(1) of this Regulation;

+ (b) any reference to a product under Regulation (EU) 2019/1020 shall be understood as including all AI systems falling within the scope of this Regulation.

+ 2. As part of their reporting obligations under Article 34(4) of Regulation (EU) 2019/1020, the market surveillance authorities shall report annually to the Commission and relevant national competition authorities any information identified in the course of market surveillance activities that may be of potential interest for the application of Union law on competition rules. They shall also annually report to the Commission about the use of prohibited practices that occurred during that year and about the measures taken.

+ 3. For high-risk AI systems related to products covered by the Union harmonisation legislation listed in Section A of Annex I, the market surveillance authority for the purposes of this Regulation shall be the authority responsible for market surveillance activities designated under those legal acts.

+ By derogation from the first subparagraph, and in appropriate circumstances, Member States may designate another relevant authority to act as a market surveillance authority, provided they ensure coordination with the relevant sectoral market surveillance authorities responsible for the enforcement of the Union harmonisation legislation listed in Annex I.

+ 4. The procedures referred to in Articles 79 to 83 of this Regulation shall not apply to AI systems related to products covered by the Union harmonisation legislation listed in section A of Annex I, where such legal acts already provide for procedures ensuring an equivalent level of protection and having the same objective. In such cases, the relevant sectoral procedures shall apply instead.

+ 5. Without prejudice to the powers of market surveillance authorities under Article 14 of Regulation (EU) 2019/1020, for the purpose of ensuring the effective enforcement of this Regulation, market surveillance authorities may exercise the powers referred to in Article 14(4), points (d) and (j), of that Regulation remotely, as appropriate.

+ 6. For high-risk AI systems placed on the market, put into service, or used by financial institutions regulated by Union financial services law, the market surveillance authority for the purposes of this Regulation shall be the relevant national authority responsible for the financial supervision of those institutions under that legislation in so far as the placing on the market, putting into service, or the use of the AI system is in direct connection with the provision of those financial services.

+ 7. By way of derogation from paragraph 6, in appropriate circumstances, and provided that coordination is ensured, another relevant authority may be identified by the Member State as market surveillance authority for the purposes of this Regulation.

+ National market surveillance authorities supervising regulated credit institutions regulated under Directive 2013/36/EU, which are participating in the Single Supervisory Mechanism established by Regulation (EU) No 1024/2013, should report, without delay, to the European Central Bank any information identified in the course of their market surveillance activities that may be of potential interest for the prudential supervisory tasks of the European Central Bank specified in that Regulation.

+ 8. For high-risk AI systems listed in point 1 of Annex III to this Regulation, in so far as the systems are used for law enforcement purposes, border management and justice and democracy, and for high-risk AI systems listed in points 6, 7 and 8 of Annex III to this Regulation, Member States shall designate as market surveillance authorities for the purposes of this Regulation either the competent data protection supervisory authorities under Regulation (EU) 2016/679 or Directive (EU) 2016/680, or any other authority designated pursuant to the same conditions laid down in Articles 41 to 44 of Directive (EU) 2016/680. Market surveillance activities shall in no way affect the independence of judicial authorities, or otherwise interfere with their activities when acting in their judicial capacity.

+ 9. Where Union institutions, bodies, offices or agencies fall within the scope of this Regulation, the European Data Protection Supervisor shall act as their market surveillance authority, except in relation to the Court of Justice of the European Union acting in its judicial capacity.

+ 10. Member States shall facilitate coordination between market surveillance authorities designated under this Regulation and other relevant national authorities or bodies which supervise the application of Union harmonisation legislation listed in Annex I, or in other Union law, that might be relevant for the high-risk AI systems referred to in Annex III.

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 103/144

+ 11. Market surveillance authorities and the Commission shall be able to propose joint activities, including joint investigations, to be conducted by either market surveillance authorities or market surveillance authorities jointly with the Commission, that have the aim of promoting compliance, identifying non-compliance, raising awareness or providing guidance in relation to this Regulation with respect to specific categories of high-risk AI systems that are found to present a serious risk across two or more Member States in accordance with Article 9 of Regulation (EU) 2019/1020. The AI Office shall provide coordination support for joint investigations.

+ 12. Without prejudice to the powers provided for under Regulation (EU) 2019/1020, and where relevant and limited to what is necessary to fulfil their tasks, the market surveillance authorities shall be granted full access by providers to the documentation as well as the training, validation and testing data sets used for the development of high-risk AI systems, including, where appropriate and subject to security safeguards, through application programming interfaces (API) or other relevant technical means and tools enabling remote access.

+ 13. Market surveillance authorities shall be granted access to the source code of the high-risk AI system upon a reasoned request and only when both of the following conditions are fulfilled:

+ (a) access to source code is necessary to assess the conformity of a high-risk AI system with the requirements set out in Chapter III, Section 2; and

+ (b) testing or auditing procedures and verifications based on the data and documentation provided by the provider have been exhausted or proved insufficient.

+ 14. Any information or documentation obtained by market surveillance authorities shall be treated in accordance with the confidentiality obligations set out in Article 78.

+ Article 75

+ Mutual assistance, market surveillance and control of general-purpose AI systems

+ 1. Where an AI system is based on a general-purpose AI model, and the model and the system are developed by the same provider, the AI Office shall have powers to monitor and supervise compliance of that AI system with obligations under this Regulation. To carry out its monitoring and supervision tasks, the AI Office shall have all the powers of a market surveillance authority provided for in this Section and Regulation (EU) 2019/1020.

+ 2. Where the relevant market surveillance authorities have sufficient reason to consider general-purpose AI systems that can be used directly by deployers for at least one purpose that is classified as high-risk pursuant to this Regulation to be non-compliant with the requirements laid down in this Regulation, they shall cooperate with the AI Office to carry out compliance evaluations, and shall inform the Board and other market surveillance authorities accordingly.

+ 3. Where a market surveillance authority is unable to conclude its investigation of the high-risk AI system because of its inability to access certain information related to the general-purpose AI model despite having made all appropriate efforts to obtain that information, it may submit a reasoned request to the AI Office, by which access to that information shall be enforced. In that case, the AI Office shall supply to the applicant authority without delay, and in any event within 30 days, any information that the AI Office considers to be relevant in order to establish whether a high-risk AI system is non-compliant. Market surveillance authorities shall safeguard the confidentiality of the information that they obtain in accordance with Article 78 of this Regulation. The procedure provided for in Chapter VI of Regulation (EU) 2019/1020 shall apply mutatis mutandis.

+ Article 76 Supervision of testing in real world conditions by market surveillance authorities

+ 1. Market surveillance authorities shall have competences and powers to ensure that testing in real world conditions is in accordance with this Regulation.

+ 104/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 2. Where testing in real world conditions is conducted for AI systems that are supervised within an AI regulatory sandbox under Article 58, the market surveillance authorities shall verify the compliance with Article 60 as part of their supervisory role for the AI regulatory sandbox. Those authorities may, as appropriate, allow the testing in real world conditions to be conducted by the provider or prospective provider, in derogation from the conditions set out in Article 60(4), points (f) and (g).

+ 3. Where a market surveillance authority has been informed by the prospective provider, the provider or any third party of a serious incident or has other grounds for considering that the conditions set out in Articles 60 and 61 are not met, it may take either of the following decisions on its territory, as appropriate:

+ (a) to suspend or terminate the testing in real world conditions;

+ (b) to require the provider or prospective provider and the deployer or prospective deployer to modify any aspect of the testing in real world conditions.

+ 4. Where a market surveillance authority has taken a decision referred to in paragraph 3 of this Article, or has issued an objection within the meaning of Article 60(4), point (b), the decision or the objection shall indicate the grounds therefor and how the provider or prospective provider can challenge the decision or objection.

+ 5. Where applicable, where a market surveillance authority has taken a decision referred to in paragraph 3, it shall communicate the grounds therefor to the market surveillance authorities of other Member States in which the AI system has been tested in accordance with the testing plan.

+ Article 77 Powers of authorities protecting fundamental rights

+ 1. National public authorities or bodies which supervise or enforce the respect of obligations under Union law protecting fundamental rights, including the right to non-discrimination, in relation to the use of high-risk AI systems referred to in Annex III shall have the power to request and access any documentation created or maintained under this Regulation in accessible language and format when access to that documentation is necessary for effectively fulfilling their mandates within the limits of their jurisdiction. The relevant public authority or body shall inform the market surveillance authority of the Member State concerned of any such request.

+ 2. By 2 November 2024, each Member State shall identify the public authorities or bodies referred to in paragraph 1 and make a list of them publicly available. Member States shall notify the list to the Commission and to the other Member States, and shall keep the list up to date.

+ 3. Where the documentation referred to in paragraph 1 is insufficient to ascertain whether an infringement of obligations under Union law protecting fundamental rights has occurred, the public authority or body referred to in paragraph 1 may make a reasoned request to the market surveillance authority, to organise testing of the high-risk AI system through technical means. The market surveillance authority shall organise the testing with the close involvement of the requesting public authority or body within a reasonable time following the request.

+ 4. Any information or documentation obtained by the national public authorities or bodies referred to in paragraph 1 of this Article pursuant to this Article shall be treated in accordance with the confidentiality obligations set out in Article 78.

+ Article 78 Confidentiality

+ 1. The Commission, market surveillance authorities and notified bodies and any other natural or legal person involved in the application of this Regulation shall, in accordance with Union or national law, respect the confidentiality of information and data obtained in carrying out their tasks and activities in such a manner as to protect, in particular:

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 105/144

+ (a) the intellectual property rights and confidential business information or trade secrets of a natural or legal person, including source code, except in the cases referred to in Article 5 of Directive (EU) 2016/943 of the European Parliament and of the Council (57);

+ (b) the effective implementation of this Regulation, in particular for the purposes of inspections, investigations or audits;

+ (c) public and national security interests;

+ (d) the conduct of criminal or administrative proceedings;

+ (e) information classified pursuant to Union or national law.

+ 2. The authorities involved in the application of this Regulation pursuant to paragraph 1 shall request only data that is strictly necessary for the assessment of the risk posed by AI systems and for the exercise of their powers in accordance with this Regulation and with Regulation (EU) 2019/1020. They shall put in place adequate and effective cybersecurity measures to protect the security and confidentiality of the information and data obtained, and shall delete the data collected as soon as it is no longer needed for the purpose for which it was obtained, in accordance with applicable Union or national law.

+ 3. Without prejudice to paragraphs 1 and 2, information exchanged on a confidential basis between the national competent authorities or between national competent authorities and the Commission shall not be disclosed without prior consultation of the originating national competent authority and the deployer when high-risk AI systems referred to in point 1, 6 or 7 of Annex III are used by law enforcement, border control, immigration or asylum authorities and when such disclosure would jeopardise public and national security interests. This exchange of information shall not cover sensitive operational data in relation to the activities of law enforcement, border control, immigration or asylum authorities.

+ When the law enforcement, immigration or asylum authorities are providers of high-risk AI systems referred to in point 1, 6 or 7 of Annex III, the technical documentation referred to in Annex IV shall remain within the premises of those authorities. Those authorities shall ensure that the market surveillance authorities referred to in Article 74(8) and (9), as applicable, can, upon request, immediately access the documentation or obtain a copy thereof. Only staff of the market surveillance authority holding the appropriate level of security clearance shall be allowed to access that documentation or any copy thereof.

+ 4. Paragraphs 1, 2 and 3 shall not affect the rights or obligations of the Commission, Member States and their relevant authorities, as well as those of notified bodies, with regard to the exchange of information and the dissemination of warnings, including in the context of cross-border cooperation, nor shall they affect the obligations of the parties concerned to provide information under criminal law of the Member States.

+ 5. The Commission and Member States may exchange, where necessary and in accordance with relevant provisions of international and trade agreements, confidential information with regulatory authorities of third countries with which they have concluded bilateral or multilateral confidentiality arrangements guaranteeing an adequate level of confidentiality.

+ Article 79 Procedure at national level for dealing with AI systems presenting a risk

+ 1. AI systems presenting a risk shall be understood as a `product presenting a risk' as defined in Article 3, point 19 of Regulation (EU) 2019/1020, in so far as they present risks to the health or safety, or to fundamental rights, of persons.

+ 2. Where the market surveillance authority of a Member State has sufficient reason to consider an AI system to present a risk as referred to in paragraph 1 of this Article, it shall carry out an evaluation of the AI system concerned in respect of its compliance with all the requirements and obligations laid down in this Regulation. Particular attention shall be given to AI systems presenting a risk to vulnerable groups. Where risks to fundamental rights are identified, the market surveillance authority shall also inform and fully cooperate with the relevant national public authorities or bodies referred to in Article 77(1). The relevant operators shall cooperate as necessary with the market surveillance authority and with the other national public authorities or bodies referred to in Article 77(1).

+ (57) Directive (EU) 2016/943 of the European Parliament and of the Council of 8 June 2016 on the protection of undisclosed know-how and business information (trade secrets) against their unlawful acquisition, use and disclosure (OJ L 157, 15.6.2016, p. 1).

+ 106/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ Where, in the course of that evaluation, the market surveillance authority or, where applicable the market surveillance authority in cooperation with the national public authority referred to in Article 77(1), finds that the AI system does not comply with the requirements and obligations laid down in this Regulation, it shall without undue delay require the relevant operator to take all appropriate corrective actions to bring the AI system into compliance, to withdraw the AI system from the market, or to recall it within a period the market surveillance authority may prescribe, and in any event within the shorter of 15 working days, or as provided for in the relevant Union harmonisation legislation.

+ The market surveillance authority shall inform the relevant notified body accordingly. Article 18 of Regulation (EU) 2019/1020 shall apply to the measures referred to in the second subparagraph of this paragraph.

+ 3. Where the market surveillance authority considers that the non-compliance is not restricted to its national territory, it shall inform the Commission and the other Member States without undue delay of the results of the evaluation and of the actions which it has required the operator to take.

+ 4. The operator shall ensure that all appropriate corrective action is taken in respect of all the AI systems concerned that it has made available on the Union market.

+ 5. Where the operator of an AI system does not take adequate corrective action within the period referred to in paragraph 2, the market surveillance authority shall take all appropriate provisional measures to prohibit or restrict the AI system's being made available on its national market or put into service, to withdraw the product or the standalone AI system from that market or to recall it. That authority shall without undue delay notify the Commission and the other Member States of those measures.

+ 6. The notification referred to in paragraph 5 shall include all available details, in particular the information necessary for the identification of the non-compliant AI system, the origin of the AI system and the supply chain, the nature of the non-compliance alleged and the risk involved, the nature and duration of the national measures taken and the arguments put forward by the relevant operator. In particular, the market surveillance authorities shall indicate whether the non-compliance is due to one or more of the following:

+ (a) non-compliance with the prohibition of the AI practices referred to in Article 5;

+ (b) a failure of a high-risk AI system to meet requirements set out in Chapter III, Section 2;

+ (c) shortcomings in the harmonised standards or common specifications referred to in Articles 40 and 41 conferring a presumption of conformity;

+ (d) non-compliance with Article 50.

+ 7. The market surveillance authorities other than the market surveillance authority of the Member State initiating the procedure shall, without undue delay, inform the Commission and the other Member States of any measures adopted and of any additional information at their disposal relating to the non-compliance of the AI system concerned, and, in the event of disagreement with the notified national measure, of their objections.

+ 8. Where, within three months of receipt of the notification referred to in paragraph 5 of this Article, no objection has been raised by either a market surveillance authority of a Member State or by the Commission in respect of a provisional measure taken by a market surveillance authority of another Member State, that measure shall be deemed justified. This shall be without prejudice to the procedural rights of the concerned operator in accordance with Article 18 of Regulation (EU) 2019/1020. The three-month period referred to in this paragraph shall be reduced to 30 days in the event of non-compliance with the prohibition of the AI practices referred to in Article 5 of this Regulation.

+ 9. The market surveillance authorities shall ensure that appropriate restrictive measures are taken in respect of the product or the AI system concerned, such as withdrawal of the product or the AI system from their market, without undue delay.

+ Article 80 Procedure for dealing with AI systems classified by the provider as non-high-risk in application of Annex III

+ 1. Where a market surveillance authority has sufficient reason to consider that an AI system classified by the provider as non-high-risk pursuant to Article 6(3) is indeed high-risk, the market surveillance authority shall carry out an evaluation of the AI system concerned in respect of its classification as a high-risk AI system based on the conditions set out in Article 6(3) and the Commission guidelines.

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 107/144

+ 2. Where, in the course of that evaluation, the market surveillance authority finds that the AI system concerned is high-risk, it shall without undue delay require the relevant provider to take all necessary actions to bring the AI system into compliance with the requirements and obligations laid down in this Regulation, as well as take appropriate corrective action within a period the market surveillance authority may prescribe.

+ 3. Where the market surveillance authority considers that the use of the AI system concerned is not restricted to its national territory, it shall inform the Commission and the other Member States without undue delay of the results of the evaluation and of the actions which it has required the provider to take.

+ 4. The provider shall ensure that all necessary action is taken to bring the AI system into compliance with the requirements and obligations laid down in this Regulation. Where the provider of an AI system concerned does not bring the AI system into compliance with those requirements and obligations within the period referred to in paragraph 2 of this Article, the provider shall be subject to fines in accordance with Article 99.

+ 5. The provider shall ensure that all appropriate corrective action is taken in respect of all the AI systems concerned that it has made available on the Union market.

+ 6. Where the provider of the AI system concerned does not take adequate corrective action within the period referred to in paragraph 2 of this Article, Article 79(5) to (9) shall apply.

+ 7. Where, in the course of the evaluation pursuant to paragraph 1 of this Article, the market surveillance authority establishes that the AI system was misclassified by the provider as non-high-risk in order to circumvent the application of requirements in Chapter III, Section 2, the provider shall be subject to fines in accordance with Article 99.

+ 8. In exercising their power to monitor the application of this Article, and in accordance with Article 11 of Regulation (EU) 2019/1020, market surveillance authorities may perform appropriate checks, taking into account in particular information stored in the EU database referred to in Article 71 of this Regulation.

+ Article 81

+ Union safeguard procedure

+ 1. Where, within three months of receipt of the notification referred to in Article 79(5), or within 30 days in the case of non-compliance with the prohibition of the AI practices referred to in Article 5, objections are raised by the market surveillance authority of a Member State to a measure taken by another market surveillance authority, or where the Commission considers the measure to be contrary to Union law, the Commission shall without undue delay enter into consultation with the market surveillance authority of the relevant Member State and the operator or operators, and shall evaluate the national measure. On the basis of the results of that evaluation, the Commission shall, within six months, or within 60 days in the case of non-compliance with the prohibition of the AI practices referred to in Article 5, starting from the notification referred to in Article 79(5), decide whether the national measure is justified and shall notify its decision to the market surveillance authority of the Member State concerned. The Commission shall also inform all other market surveillance authorities of its decision.

+ 2. Where the Commission considers the measure taken by the relevant Member State to be justified, all Member States shall ensure that they take appropriate restrictive measures in respect of the AI system concerned, such as requiring the withdrawal of the AI system from their market without undue delay, and shall inform the Commission accordingly. Where the Commission considers the national measure to be unjustified, the Member State concerned shall withdraw the measure and shall inform the Commission accordingly.

+ 3. Where the national measure is considered justified and the non-compliance of the AI system is attributed to shortcomings in the harmonised standards or common specifications referred to in Articles 40 and 41 of this Regulation, the Commission shall apply the procedure provided for in Article 11 of Regulation (EU) No 1025/2012.

+ Article 82 Compliant AI systems which present a risk

+ 1. Where, having performed an evaluation under Article 79, after consulting the relevant national public authority referred to in Article 77(1), the market surveillance authority of a Member State finds that although a high-risk AI system complies with this Regulation, it nevertheless presents a risk to the health or safety of persons, to fundamental rights, or to other aspects of public interest protection, it shall require the relevant operator to take all appropriate measures to ensure that the AI system concerned, when placed on the market or put into service, no longer presents that risk without undue delay, within a period it may prescribe.

+ 108/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 2. The provider or other relevant operator shall ensure that corrective action is taken in respect of all the AI systems concerned that it has made available on the Union market within the timeline prescribed by the market surveillance authority of the Member State referred to in paragraph 1.

+ 3. The Member States shall immediately inform the Commission and the other Member States of a finding under paragraph 1. That information shall include all available details, in particular the data necessary for the identification of the AI system concerned, the origin and the supply chain of the AI system, the nature of the risk involved and the nature and duration of the national measures taken.

+ 4. The Commission shall without undue delay enter into consultation with the Member States concerned and the relevant operators, and shall evaluate the national measures taken. On the basis of the results of that evaluation, the Commission shall decide whether the measure is justified and, where necessary, propose other appropriate measures.

+ 5. The Commission shall immediately communicate its decision to the Member States concerned and to the relevant operators. It shall also inform the other Member States.

+ Article 83 Formal non-compliance

+ 1. Where the market surveillance authority of a Member State makes one of the following findings, it shall require the relevant provider to put an end to the non-compliance concerned, within a period it may prescribe:

+ (a) the CE marking has been affixed in violation of Article 48;

+ (b) the CE marking has not been affixed;

+ (c) the EU declaration of conformity referred to in Article 47 has not been drawn up;

+ (d) the EU declaration of conformity referred to in Article 47 has not been drawn up correctly;

+ (e) the registration in the EU database referred to in Article 71 has not been carried out;

+ (f) where applicable, no authorised representative has been appointed;

+ (g) technical documentation is not available.

+ 2. Where the non-compliance referred to in paragraph 1 persists, the market surveillance authority of the Member State concerned shall take appropriate and proportionate measures to restrict or prohibit the high-risk AI system being made available on the market or to ensure that it is recalled or withdrawn from the market without delay.

+ Article 84 Union AI testing support structures

+ 1. The Commission shall designate one or more Union AI testing support structures to perform the tasks listed under Article 21(6) of Regulation (EU) 2019/1020 in the area of AI.

+ 2. Without prejudice to the tasks referred to in paragraph 1, Union AI testing support structures shall also provide independent technical or scientific advice at the request of the Board, the Commission, or of market surveillance authorities.

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 109/144

+ SECTION 4

+ Remedies

+ Article 85 Right to lodge a complaint with a market surveillance authority

+ Without prejudice to other administrative or judicial remedies, any natural or legal person having grounds to consider that there has been an infringement of the provisions of this Regulation may submit complaints to the relevant market surveillance authority.

+ In accordance with Regulation (EU) 2019/1020, such complaints shall be taken into account for the purpose of conducting market surveillance activities, and shall be handled in line with the dedicated procedures established therefor by the market surveillance authorities.

+ Article 86 Right to explanation of individual decision-making

+ 1. Any affected person subject to a decision which is taken by the deployer on the basis of the output from a high-risk AI system listed in Annex III, with the exception of systems listed under point 2 thereof, and which produces legal effects or similarly significantly affects that person in a way that they consider to have an adverse impact on their health, safety or fundamental rights shall have the right to obtain from the deployer clear and meaningful explanations of the role of the AI system in the decision-making procedure and the main elements of the decision taken.

+ 2. Paragraph 1 shall not apply to the use of AI systems for which exceptions from, or restrictions to, the obligation under that paragraph follow from Union or national law in compliance with Union law.

+ 3. This Article shall apply only to the extent that the right referred to in paragraph 1 is not otherwise provided for under Union law.

+ Article 87 Reporting of infringements and protection of reporting persons

+ Directive (EU) 2019/1937 shall apply to the reporting of infringements of this Regulation and the protection of persons reporting such infringements.

+ SECTION 5

+ Supervision, investigation, enforcement and monitoring in respect of providers of general-purpose AI models

+ Article 88 Enforcement of the obligations of providers of general-purpose AI models

+ 1. The Commission shall have exclusive powers to supervise and enforce Chapter V, taking into account the procedural guarantees under Article 94. The Commission shall entrust the implementation of these tasks to the AI Office, without prejudice to the powers of organisation of the Commission and the division of competences between Member States and the Union based on the Treaties.

+ 2. Without prejudice to Article 75(3), market surveillance authorities may request the Commission to exercise the powers laid down in this Section, where that is necessary and proportionate to assist with the fulfilment of their tasks under this Regulation.

+ 110/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ Article 89 Monitoring actions

+ 1. For the purpose of carrying out the tasks assigned to it under this Section, the AI Office may take the necessary actions to monitor the effective implementation and compliance with this Regulation by providers of general-purpose AI models, including their adherence to approved codes of practice.

+ 2. Downstream providers shall have the right to lodge a complaint alleging an infringement of this Regulation. A complaint shall be duly reasoned and indicate at least:

+ (a) the point of contact of the provider of the general-purpose AI model concerned;

+ (b) a description of the relevant facts, the provisions of this Regulation concerned, and the reason why the downstream provider considers that the provider of the general-purpose AI model concerned infringed this Regulation;

+ (c) any other information that the downstream provider that sent the request considers relevant, including, where appropriate, information gathered on its own initiative.

+ Article 90 Alerts of systemic risks by the scientific panel

+ 1. The scientific panel may provide a qualified alert to the AI Office where it has reason to suspect that:

+ (a) a general-purpose AI model poses concrete identifiable risk at Union level; or

+ (b) a general-purpose AI model meets the conditions referred to in Article 51.

+ 2. Upon such qualified alert, the Commission, through the AI Office and after having informed the Board, may exercise the powers laid down in this Section for the purpose of assessing the matter. The AI Office shall inform the Board of any measure according to Articles 91 to 94.

+ 3. A qualified alert shall be duly reasoned and indicate at least:

+ (a) the point of contact of the provider of the general-purpose AI model with systemic risk concerned;

+ (b) a description of the relevant facts and the reasons for the alert by the scientific panel;

+ (c) any other information that the scientific panel considers to be relevant, including, where appropriate, information gathered on its own initiative.

+ Article 91 Power to request documentation and information

+ 1. The Commission may request the provider of the general-purpose AI model concerned to provide the documentation drawn up by the provider in accordance with Articles 53 and 55, or any additional information that is necessary for the purpose of assessing compliance of the provider with this Regulation.

+ 2. Before sending the request for information, the AI Office may initiate a structured dialogue with the provider of the general-purpose AI model.

+ 3. Upon a duly substantiated request from the scientific panel, the Commission may issue a request for information to a provider of a general-purpose AI model, where the access to information is necessary and proportionate for the fulfilment of the tasks of the scientific panel under Article 68(2).

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 111/144

+ 4. The request for information shall state the legal basis and the purpose of the request, specify what information is required, set a period within which the information is to be provided, and indicate the fines provided for in Article 101 for supplying incorrect, incomplete or misleading information.

+ 5. The provider of the general-purpose AI model concerned, or its representative shall supply the information requested. In the case of legal persons, companies or firms, or where the provider has no legal personality, the persons authorised to represent them by law or by their statutes, shall supply the information requested on behalf of the provider of the general-purpose AI model concerned. Lawyers duly authorised to act may supply information on behalf of their clients. The clients shall nevertheless remain fully responsible if the information supplied is incomplete, incorrect or misleading.

+ Article 92 Power to conduct evaluations

+ 1. The AI Office, after consulting the Board, may conduct evaluations of the general-purpose AI model concerned:

+ (a) to assess compliance of the provider with obligations under this Regulation, where the information gathered pursuant to Article 91 is insufficient; or

+ (b) to investigate systemic risks at Union level of general-purpose AI models with systemic risk, in particular following a qualified alert from the scientific panel in accordance with Article 90(1), point (a).

+ 2. The Commission may decide to appoint independent experts to carry out evaluations on its behalf, including from the scientific panel established pursuant to Article 68. Independent experts appointed for this task shall meet the criteria outlined in Article 68(2).

+ 3. For the purposes of paragraph 1, the Commission may request access to the general-purpose AI model concerned through APIs or further appropriate technical means and tools, including source code.

+ 4. The request for access shall state the legal basis, the purpose and reasons of the request and set the period within which the access is to be provided, and the fines provided for in Article 101 for failure to provide access.

+ 5. The providers of the general-purpose AI model concerned or its representative shall supply the information requested. In the case of legal persons, companies or firms, or where the provider has no legal personality, the persons authorised to represent them by law or by their statutes, shall provide the access requested on behalf of the provider of the general-purpose AI model concerned.

+ 6. The Commission shall adopt implementing acts setting out the detailed arrangements and the conditions for the evaluations, including the detailed arrangements for involving independent experts, and the procedure for the selection thereof. Those implementing acts shall be adopted in accordance with the examination procedure referred to in Article 98(2).

+ 7. Prior to requesting access to the general-purpose AI model concerned, the AI Office may initiate a structured dialogue with the provider of the general-purpose AI model to gather more information on the internal testing of the model, internal safeguards for preventing systemic risks, and other internal procedures and measures the provider has taken to mitigate such risks.

+ Article 93 Power to request measures

+ 1. Where necessary and appropriate, the Commission may request providers to:

+ (a) take appropriate measures to comply with the obligations set out in Articles 53 and 54;

+ 112/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ (b) implement mitigation measures, where the evaluation carried out in accordance with Article 92 has given rise to serious and substantiated concern of a systemic risk at Union level;

+ (c) restrict the making available on the market, withdraw or recall the model.

+ 2. Before a measure is requested, the AI Office may initiate a structured dialogue with the provider of the general-purpose AI model.

+ 3. If, during the structured dialogue referred to in paragraph 2, the provider of the general-purpose AI model with systemic risk offers commitments to implement mitigation measures to address a systemic risk at Union level, the Commission may, by decision, make those commitments binding and declare that there are no further grounds for action.

+ Article 94 Procedural rights of economic operators of the general-purpose AI model

+ Article 18 of Regulation (EU) 2019/1020 shall apply mutatis mutandis to the providers of the general-purpose AI model, without prejudice to more specific procedural rights provided for in this Regulation.

+ CHAPTER X CODES OF CONDUCT AND GUIDELINES

+ Article 95 Codes of conduct for voluntary application of specific requirements

+ 1. The AI Office and the Member States shall encourage and facilitate the drawing up of codes of conduct, including related governance mechanisms, intended to foster the voluntary application to AI systems, other than high-risk AI systems, of some or all of the requirements set out in Chapter III, Section 2 taking into account the available technical solutions and industry best practices allowing for the application of such requirements.

+ 2. The AI Office and the Member States shall facilitate the drawing up of codes of conduct concerning the voluntary application, including by deployers, of specific requirements to all AI systems, on the basis of clear objectives and key performance indicators to measure the achievement of those objectives, including elements such as, but not limited to:

+ (a) applicable elements provided for in Union ethical guidelines for trustworthy AI;

+ (b) assessing and minimising the impact of AI systems on environmental sustainability, including as regards energy-efficient programming and techniques for the efficient design, training and use of AI;

+ (c) promoting AI literacy, in particular that of persons dealing with the development, operation and use of AI;

+ (d) facilitating an inclusive and diverse design of AI systems, including through the establishment of inclusive and diverse development teams and the promotion of stakeholders' participation in that process;

+ (e) assessing and preventing the negative impact of AI systems on vulnerable persons or groups of vulnerable persons, including as regards accessibility for persons with a disability, as well as on gender equality.

+ 3. Codes of conduct may be drawn up by individual providers or deployers of AI systems or by organisations representing them or by both, including with the involvement of any interested stakeholders and their representative organisations, including civil society organisations and academia. Codes of conduct may cover one or more AI systems taking into account the similarity of the intended purpose of the relevant systems.

+ 4. The AI Office and the Member States shall take into account the specific interests and needs of SMEs, including start-ups, when encouraging and facilitating the drawing up of codes of conduct.

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 113/144

+ Article 96 Guidelines from the Commission on the implementation of this Regulation

+ 1. The Commission shall develop guidelines on the practical implementation of this Regulation, and in particular on:

+ (a) the application of the requirements and obligations referred to in Articles 8 to 15 and in Article 25;

+ (b) the prohibited practices referred to in Article 5;

+ (c) the practical implementation of the provisions related to substantial modification;

+ (d) the practical implementation of transparency obligations laid down in Article 50;

+ (e) detailed information on the relationship of this Regulation with the Union harmonisation legislation listed in Annex I, as well as with other relevant Union law, including as regards consistency in their enforcement;

+ (f) the application of the definition of an AI system as set out in Article 3, point (1).

+ When issuing such guidelines, the Commission shall pay particular attention to the needs of SMEs including start-ups, of local public authorities and of the sectors most likely to be affected by this Regulation.

+ The guidelines referred to in the first subparagraph of this paragraph shall take due account of the generally acknowledged state of the art on AI, as well as of relevant harmonised standards and common specifications that are referred to in Articles 40 and 41, or of those harmonised standards or technical specifications that are set out pursuant to Union harmonisation law.

+ 2. At the request of the Member States or the AI Office, or on its own initiative, the Commission shall update guidelines previously adopted when deemed necessary.

+ CHAPTER XI DELEGATION OF POWER AND COMMITTEE PROCEDURE

+ Article 97 Exercise of the delegation

+ 1. The power to adopt delegated acts is conferred on the Commission subject to the conditions laid down in this Article.

+ 2. The power to adopt delegated acts referred to in Article 6(6) and (7), Article 7(1) and (3), Article 11(3), Article 43(5) and (6), Article 47(5), Article 51(3), Article 52(4) and Article 53(5) and (6) shall be conferred on the Commission for a period of five years from 1 August 2024. The Commission shall draw up a report in respect of the delegation of power not later than nine months before the end of the five-year period. The delegation of power shall be tacitly extended for periods of an identical duration, unless the European Parliament or the Council opposes such extension not later than three months before the end of each period.

+ 3. The delegation of power referred to in Article 6(6) and (7), Article 7(1) and (3), Article 11(3), Article 43(5) and (6), Article 47(5), Article 51(3), Article 52(4) and Article 53(5) and (6) may be revoked at any time by the European Parliament or by the Council. A decision of revocation shall put an end to the delegation of power specified in that decision. It shall take effect the day following that of its publication in the Official Journal of the European Union or at a later date specified therein. It shall not affect the validity of any delegated acts already in force.

+ 4. Before adopting a delegated act, the Commission shall consult experts designated by each Member State in accordance with the principles laid down in the Interinstitutional Agreement of 13 April 2016 on Better Law-Making.

+ 114/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 5. As soon as it adopts a delegated act, the Commission shall notify it simultaneously to the European Parliament and to the Council.

+ 6. Any delegated act adopted pursuant to Article 6(6) or (7), Article 7(1) or (3), Article 11(3), Article 43(5) or (6), Article 47(5), Article 51(3), Article 52(4) or Article 53(5) or (6) shall enter into force only if no objection has been expressed by either the European Parliament or the Council within a period of three months of notification of that act to the European Parliament and the Council or if, before the expiry of that period, the European Parliament and the Council have both informed the Commission that they will not object. That period shall be extended by three months at the initiative of the European Parliament or of the Council.

+ Article 98 Committee procedure

+ 1. The Commission shall be assisted by a committee. That committee shall be a committee within the meaning of Regulation (EU) No 182/2011.

+ 2. Where reference is made to this paragraph, Article 5 of Regulation (EU) No 182/2011 shall apply.

+ CHAPTER XII PENALTIES

+ Article 99 Penalties

+ 1. In accordance with the terms and conditions laid down in this Regulation, Member States shall lay down the rules on penalties and other enforcement measures, which may also include warnings and non-monetary measures, applicable to infringements of this Regulation by operators, and shall take all measures necessary to ensure that they are properly and effectively implemented, thereby taking into account the guidelines issued by the Commission pursuant to Article 96. The penalties provided for shall be effective, proportionate and dissuasive. They shall take into account the interests of SMEs, including start-ups, and their economic viability.

+ 2. The Member States shall, without delay and at the latest by the date of entry into application, notify the Commission of the rules on penalties and of other enforcement measures referred to in paragraph 1, and shall notify it, without delay, of any subsequent amendment to them.

+ 3. Non-compliance with the prohibition of the AI practices referred to in Article 5 shall be subject to administrative fines of up to EUR 35 000 000 or, if the offender is an undertaking, up to 7 % of its total worldwide annual turnover for the preceding financial year, whichever is higher.

+ 4. Non-compliance with any of the following provisions related to operators or notified bodies, other than those laid down in Articles 5, shall be subject to administrative fines of up to EUR 15 000 000 or, if the offender is an undertaking, up to 3 % of its total worldwide annual turnover for the preceding financial year, whichever is higher:

+ (a) obligations of providers pursuant to Article 16;

+ (b) obligations of authorised representatives pursuant to Article 22;

+ (c) obligations of importers pursuant to Article 23;

+ (d) obligations of distributors pursuant to Article 24;

+ (e) obligations of deployers pursuant to Article 26;

+ (f) requirements and obligations of notified bodies pursuant to Article 31, Article 33(1), (3) and (4) or Article 34;

+ (g) transparency obligations for providers and deployers pursuant to Article 50. ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 115/144

+ 5. The supply of incorrect, incomplete or misleading information to notified bodies or national competent authorities in reply to a request shall be subject to administrative fines of up to EUR 7 500 000 or, if the offender is an undertaking, up to 1 % of its total worldwide annual turnover for the preceding financial year, whichever is higher.

+ 6. In the case of SMEs, including start-ups, each fine referred to in this Article shall be up to the percentages or amount referred to in paragraphs 3, 4 and 5, whichever thereof is lower.

+ 7. When deciding whether to impose an administrative fine and when deciding on the amount of the administrative fine in each individual case, all relevant circumstances of the specific situation shall be taken into account and, as appropriate, regard shall be given to the following:

+ (a) the nature, gravity and duration of the infringement and of its consequences, taking into account the purpose of the AI system, as well as, where appropriate, the number of affected persons and the level of damage suffered by them;

+ (b) whether administrative fines have already been applied by other market surveillance authorities to the same operator for the same infringement;

+ (c) whether administrative fines have already been applied by other authorities to the same operator for infringements of other Union or national law, when such infringements result from the same activity or omission constituting a relevant infringement of this Regulation;

+ (d) the size, the annual turnover and market share of the operator committing the infringement;

+ (e) any other aggravating or mitigating factor applicable to the circumstances of the case, such as financial benefits gained, or losses avoided, directly or indirectly, from the infringement;

+ (f) the degree of cooperation with the national competent authorities, in order to remedy the infringement and mitigate the possible adverse effects of the infringement;

+ (g) the degree of responsibility of the operator taking into account the technical and organisational measures implemented by it;

+ (h) the manner in which the infringement became known to the national competent authorities, in particular whether, and if so to what extent, the operator notified the infringement;

+ (i) the intentional or negligent character of the infringement;

+ (j) any action taken by the operator to mitigate the harm suffered by the affected persons.

+ 8. Each Member State shall lay down rules on to what extent administrative fines may be imposed on public authorities and bodies established in that Member State.

+ 9. Depending on the legal system of the Member States, the rules on administrative fines may be applied in such a manner that the fines are imposed by competent national courts or by other bodies, as applicable in those Member States. The application of such rules in those Member States shall have an equivalent effect.

+ 10. The exercise of powers under this Article shall be subject to appropriate procedural safeguards in accordance with Union and national law, including effective judicial remedies and due process.

+ 11. Member States shall, on an annual basis, report to the Commission about the administrative fines they have issued during that year, in accordance with this Article, and about any related litigation or judicial proceedings.

+ Article 100 Administrative fines on Union institutions, bodies, offices and agencies

+ 1. The European Data Protection Supervisor may impose administrative fines on Union institutions, bodies, offices and agencies falling within the scope of this Regulation. When deciding whether to impose an administrative fine and when deciding on the amount of the administrative fine in each individual case, all relevant circumstances of the specific situation shall be taken into account and due regard shall be given to the following:

+ 116/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ (a) the nature, gravity and duration of the infringement and of its consequences, taking into account the purpose of the AI system concerned, as well as, where appropriate, the number of affected persons and the level of damage suffered by them;

+ (b) the degree of responsibility of the Union institution, body, office or agency, taking into account technical and organisational measures implemented by them;

+ (c) any action taken by the Union institution, body, office or agency to mitigate the damage suffered by affected persons;

+ (d) the degree of cooperation with the European Data Protection Supervisor in order to remedy the infringement and mitigate the possible adverse effects of the infringement, including compliance with any of the measures previously ordered by the European Data Protection Supervisor against the Union institution, body, office or agency concerned with regard to the same subject matter;

+ (e) any similar previous infringements by the Union institution, body, office or agency;

+ (f) the manner in which the infringement became known to the European Data Protection Supervisor, in particular whether, and if so to what extent, the Union institution, body, office or agency notified the infringement;

+ (g) the annual budget of the Union institution, body, office or agency.

+ 2. Non-compliance with the prohibition of the AI practices referred to in Article 5 shall be subject to administrative fines of up to EUR 1 500 000.

+ 3. The non-compliance of the AI system with any requirements or obligations under this Regulation, other than those laid down in Article 5, shall be subject to administrative fines of up to EUR 750 000.

+ 4. Before taking decisions pursuant to this Article, the European Data Protection Supervisor shall give the Union institution, body, office or agency which is the subject of the proceedings conducted by the European Data Protection Supervisor the opportunity of being heard on the matter regarding the possible infringement. The European Data Protection Supervisor shall base his or her decisions only on elements and circumstances on which the parties concerned have been able to comment. Complainants, if any, shall be associated closely with the proceedings.

+ 5. The rights of defence of the parties concerned shall be fully respected in the proceedings. They shall be entitled to have access to the European Data Protection Supervisor's file, subject to the legitimate interest of individuals or undertakings in the protection of their personal data or business secrets.

+ 6. Funds collected by imposition of fines in this Article shall contribute to the general budget of the Union. The fines shall not affect the effective operation of the Union institution, body, office or agency fined.

+ 7. The European Data Protection Supervisor shall, on an annual basis, notify the Commission of the administrative fines it has imposed pursuant to this Article and of any litigation or judicial proceedings it has initiated.

+ Article 101 Fines for providers of general-purpose AI models

+ 1. The Commission may impose on providers of general-purpose AI models fines not exceeding 3 % of their annual total worldwide turnover in the preceding financial year or EUR 15 000 000, whichever is higher., when the Commission finds that the provider intentionally or negligently:

+ (a) infringed the relevant provisions of this Regulation;

+ (b) failed to comply with a request for a document or for information pursuant to Article 91, or supplied incorrect, incomplete or misleading information;

+ (c) failed to comply with a measure requested under Article 93; ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 117/144

+ (d) failed to make available to the Commission access to the general-purpose AI model or general-purpose AI model with systemic risk with a view to conducting an evaluation pursuant to Article 92.

+ In fixing the amount of the fine or periodic penalty payment, regard shall be had to the nature, gravity and duration of the infringement, taking due account of the principles of proportionality and appropriateness. The Commission shall also into account commitments made in accordance with Article 93(3) or made in relevant codes of practice in accordance with Article 56.

+ 2. Before adopting the decision pursuant to paragraph 1, the Commission shall communicate its preliminary findings to the provider of the general-purpose AI model and give it an opportunity to be heard.

+ 3. Fines imposed in accordance with this Article shall be effective, proportionate and dissuasive.

+ 4. Information on fines imposed under this Article shall also be communicated to the Board as appropriate.

+ 5. The Court of Justice of the European Union shall have unlimited jurisdiction to review decisions of the Commission fixing a fine under this Article. It may cancel, reduce or increase the fine imposed.

+ 6. The Commission shall adopt implementing acts containing detailed arrangements and procedural safeguards for proceedings in view of the possible adoption of decisions pursuant to paragraph 1 of this Article. Those implementing acts shall be adopted in accordance with the examination procedure referred to in Article 98(2).

+ CHAPTER XIII FINAL PROVISIONS

+ Article 102

+ Amendment to Regulation (EC) No 300/2008

+ In Article 4(3) of Regulation (EC) No 300/2008, the following subparagraph is added:

+ `When adopting detailed measures related to technical specifications and procedures for approval and use of security equipment concerning Artificial Intelligence systems within the meaning of Regulation (EU) 2024/1689 of the European Parliament and of the Council (*), the requirements set out in Chapter III, Section 2, of that Regulation shall be taken into account.

+ (*) Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 laying down harmonised rules on artificial intelligence and amending Regulations (EC) No 300/2008, (EU) No 167/2013, (EU) No 168/2013, (EU) 2018/858, (EU) 2018/1139 and (EU) 2019/2144 and Directives 2014/90/EU, (EU) 2016/797 and (EU) 2020/1828 (Artificial Intelligence Act) (OJ L, 2024/1689, 12.7.2024, ELI: http://data.europa.eu/eli/reg/ 2024/1689/oj).'.

+ Article 103 Amendment to Regulation (EU) No 167/2013

+ In Article 17(5) of Regulation (EU) No 167/2013, the following subparagraph is added:

+ `When adopting delegated acts pursuant to the first subparagraph concerning artificial intelligence systems which are safety components within the meaning of Regulation (EU) 2024/1689 of the European Parliament and of the Council (*), the requirements set out in Chapter III, Section 2, of that Regulation shall be taken into account.

+ 118/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ Article 104 Amendment to Regulation (EU) No 168/2013

+ In Article 22(5) of Regulation (EU) No 168/2013, the following subparagraph is added:

+ `When adopting delegated acts pursuant to the first subparagraph concerning Artificial Intelligence systems which are safety components within the meaning of Regulation (EU) 2024/1689 of the European Parliament and of the Council (*), the requirements set out in Chapter III, Section 2, of that Regulation shall be taken into account.

+ Article 105 Amendment to Directive 2014/90/EU

+ In Article 8 of Directive 2014/90/EU, the following paragraph is added:

+ `5. For Artificial Intelligence systems which are safety components within the meaning of Regulation (EU) 2024/1689 of the European Parliament and of the Council (*), when carrying out its activities pursuant to paragraph 1 and when adopting technical specifications and testing standards in accordance with paragraphs 2 and 3, the Commission shall take into account the requirements set out in Chapter III, Section 2, of that Regulation.

+ Article 106 Amendment to Directive (EU) 2016/797

+ In Article 5 of Directive (EU) 2016/797, the following paragraph is added:

+ `12. When adopting delegated acts pursuant to paragraph 1 and implementing acts pursuant to paragraph 11 concerning Artificial Intelligence systems which are safety components within the meaning of Regulation (EU) 2024/1689 of the European Parliament and of the Council (*), the requirements set out in Chapter III, Section 2, of that Regulation shall be taken into account.

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 119/144

+ Article 107 Amendment to Regulation (EU) 2018/858

+ In Article 5 of Regulation (EU) 2018/858 the following paragraph is added:

+ `4. When adopting delegated acts pursuant to paragraph 3 concerning Artificial Intelligence systems which are safety components within the meaning of Regulation (EU) 2024/1689 of the European Parliament and of the Council (*), the requirements set out in Chapter III, Section 2, of that Regulation shall be taken into account.

+ Article 108 Amendments to Regulation (EU) 2018/1139

+ Regulation (EU) 2018/1139 is amended as follows:

+ (1) in Article 17, the following paragraph is added:

+ `3. Without prejudice to paragraph 2, when adopting implementing acts pursuant to paragraph 1 concerning Artificial Intelligence systems which are safety components within the meaning of Regulation (EU) 2024/1689 of the European Parliament and of the Council (*), the requirements set out in Chapter III, Section 2, of that Regulation shall be taken into account.

+ (2) in Article 19, the following paragraph is added:

+ `4. When adopting delegated acts pursuant to paragraphs 1 and 2 concerning Artificial Intelligence systems which are safety components within the meaning of Regulation (EU) 2024/1689, the requirements set out in Chapter III, Section 2, of that Regulation shall be taken into account.';

+ (3) in Article 43, the following paragraph is added:

+ `4. When adopting implementing acts pursuant to paragraph 1 concerning Artificial Intelligence systems which are safety components within the meaning of Regulation (EU) 2024/1689, the requirements set out in Chapter III, Section 2, of that Regulation shall be taken into account.';

+ (4) in Article 47, the following paragraph is added:

+ `3. When adopting delegated acts pursuant to paragraphs 1 and 2 concerning Artificial Intelligence systems which are safety components within the meaning of Regulation (EU) 2024/1689, the requirements set out in Chapter III, Section 2, of that Regulation shall be taken into account.';

+ (5) in Article 57, the following subparagraph is added:

+ `When adopting those implementing acts concerning Artificial Intelligence systems which are safety components within the meaning of Regulation (EU) 2024/1689, the requirements set out in Chapter III, Section 2, of that Regulation shall be taken into account.';

+ 120/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ (6) in Article 58, the following paragraph is added:

+ Article 109 Amendment to Regulation (EU) 2019/2144

+ In Article 11 of Regulation (EU) 2019/2144, the following paragraph is added:

+ `3. When adopting the implementing acts pursuant to paragraph 2, concerning artificial intelligence systems which are safety components within the meaning of Regulation (EU) 2024/1689 of the European Parliament and of the Council (*), the requirements set out in Chapter III, Section 2, of that Regulation shall be taken into account.

+ Article 110 Amendment to Directive (EU) 2020/1828

+ In Annex I to Directive (EU) 2020/1828 of the European Parliament and of the Council (58), the following point is added:

+ `(68) Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 laying down harmonised rules on artificial intelligence and amending Regulations (EC) No 300/2008, (EU) No 167/2013, (EU) No 168/2013, (EU) 2018/858, (EU) 2018/1139 and (EU) 2019/2144 and Directives 2014/90/EU, (EU) 2016/797 and (EU) 2020/1828 (Artificial Intelligence Act) (OJ L, 2024/1689, 12.7.2024, ELI: http://data.europa.eu/eli/reg/ 2024/1689/oj).'.

+ Article 111

+ AI systems already placed on the market or put into service and general-purpose AI models already placed on the marked

+ 1. Without prejudice to the application of Article 5 as referred to in Article 113(3), point (a), AI systems which are components of the large-scale IT systems established by the legal acts listed in Annex X that have been placed on the market or put into service before 2 August 2027 shall be brought into compliance with this Regulation by 31 December 2030.

+ The requirements laid down in this Regulation shall be taken into account in the evaluation of each large-scale IT system established by the legal acts listed in Annex X to be undertaken as provided for in those legal acts and where those legal acts are replaced or amended.

+ 2. Without prejudice to the application of Article 5 as referred to in Article 113(3), point (a), this Regulation shall apply to operators of high-risk AI systems, other than the systems referred to in paragraph 1 of this Article, that have been placed on the market or put into service before 2 August 2026, only if, as from that date, those systems are subject to significant changes in their designs. In any case, the providers and deployers of high-risk AI systems intended to be used by public authorities shall take the necessary steps to comply with the requirements and obligations of this Regulation by 2 August 2030.

+ 3. Providers of general-purpose AI models that have been placed on the market before 2 August 2025 shall take the necessary steps in order to comply with the obligations laid down in this Regulation by 2 August 2027.

+ (58) Directive (EU) 2020/1828 of the European Parliament and of the Council of 25 November 2020 on representative actions for the protection of the collective interests of consumers and repealing Directive 2009/22/EC (OJ L 409, 4.12.2020, p. 1).

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 121/144

+ Article 112 Evaluation and review

+ 1. The Commission shall assess the need for amendment of the list set out in Annex III and of the list of prohibited AI practices laid down in Article 5, once a year following the entry into force of this Regulation, and until the end of the period of the delegation of power laid down in Article 97. The Commission shall submit the findings of that assessment to the European Parliament and the Council.

+ 2. By 2 August 2028 and every four years thereafter, the Commission shall evaluate and report to the European Parliament and to the Council on the following:

+ (a) the need for amendments extending existing area headings or adding new area headings in Annex III;

+ (b) amendments to the list of AI systems requiring additional transparency measures in Article 50;

+ (c) amendments enhancing the effectiveness of the supervision and governance system.

+ 3. By 2 August 2029 and every four years thereafter, the Commission shall submit a report on the evaluation and review of this Regulation to the European Parliament and to the Council. The report shall include an assessment with regard to the structure of enforcement and the possible need for a Union agency to resolve any identified shortcomings. On the basis of the findings, that report shall, where appropriate, be accompanied by a proposal for amendment of this Regulation. The reports shall be made public.

+ 4. The reports referred to in paragraph 2 shall pay specific attention to the following:

+ (a) the status of the financial, technical and human resources of the national competent authorities in order to effectively perform the tasks assigned to them under this Regulation;

+ (b) the state of penalties, in particular administrative fines as referred to in Article 99(1), applied by Member States for infringements of this Regulation;

+ (c) adopted harmonised standards and common specifications developed to support this Regulation;

+ (d) the number of undertakings that enter the market after the entry into application of this Regulation, and how many of them are SMEs.

+ 5. By 2 August 2028, the Commission shall evaluate the functioning of the AI Office, whether the AI Office has been given sufficient powers and competences to fulfil its tasks, and whether it would be relevant and needed for the proper implementation and enforcement of this Regulation to upgrade the AI Office and its enforcement competences and to increase its resources. The Commission shall submit a report on its evaluation to the European Parliament and to the Council.

+ 6. By 2 August 2028 and every four years thereafter, the Commission shall submit a report on the review of the progress on the development of standardisation deliverables on the energy-efficient development of general-purpose AI models, and asses the need for further measures or actions, including binding measures or actions. The report shall be submitted to the European Parliament and to the Council, and it shall be made public.

+ 7. By 2 August 2028 and every three years thereafter, the Commission shall evaluate the impact and effectiveness of voluntary codes of conduct to foster the application of the requirements set out in Chapter III, Section 2 for AI systems other than high-risk AI systems and possibly other additional requirements for AI systems other than high-risk AI systems, including as regards environmental sustainability.

+ 8. For the purposes of paragraphs 1 to 7, the Board, the Member States and national competent authorities shall provide the Commission with information upon its request and without undue delay.

+ 9. In carrying out the evaluations and reviews referred to in paragraphs 1 to 7, the Commission shall take into account the positions and findings of the Board, of the European Parliament, of the Council, and of other relevant bodies or sources.

+ 122/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 10. The Commission shall, if necessary, submit appropriate proposals to amend this Regulation, in particular taking into account developments in technology, the effect of AI systems on health and safety, and on fundamental rights, and in light of the state of progress in the information society.

+ 11. To guide the evaluations and reviews referred to in paragraphs 1 to 7 of this Article, the AI Office shall undertake to develop an objective and participative methodology for the evaluation of risk levels based on the criteria outlined in the relevant Articles and the inclusion of new systems in:

+ (a) the list set out in Annex III, including the extension of existing area headings or the addition of new area headings in that Annex;

+ (b) the list of prohibited practices set out in Article 5; and

+ (c) the list of AI systems requiring additional transparency measures pursuant to Article 50.

+ 12. Any amendment to this Regulation pursuant to paragraph 10, or relevant delegated or implementing acts, which concerns sectoral Union harmonisation legislation listed in Section B of Annex I shall take into account the regulatory specificities of each sector, and the existing governance, conformity assessment and enforcement mechanisms and authorities established therein.

+ 13. By 2 August 2031, the Commission shall carry out an assessment of the enforcement of this Regulation and shall report on it to the European Parliament, the Council and the European Economic and Social Committee, taking into account the first years of application of this Regulation. On the basis of the findings, that report shall, where appropriate, be accompanied by a proposal for amendment of this Regulation with regard to the structure of enforcement and the need for a Union agency to resolve any identified shortcomings.

+ Article 113 Entry into force and application This Regulation shall enter into force on the twentieth day following that of its publication in the Official Journal of the European Union. It shall apply from 2 August 2026. However: (a) Chapters I and II shall apply from 2 February 2025; (b) Chapter III Section 4, Chapter V, Chapter VII and Chapter XII and Article 78 shall apply from 2 August 2025, with the exception of Article 101; (c) Article 6(1) and the corresponding obligations in this Regulation shall apply from 2 August 2027.

+ This Regulation shall be binding in its entirety and directly applicable in all Member States.

+ Done at Brussels, 13 June 2024.

+ For the European Parliament The President R. METSOLA

+ For the Council The President M. MICHEL

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 123/144

+ ANNEX I

+ List of Union harmonisation legislation

+ Section A. List of Union harmonisation legislation based on the New Legislative Framework

+ 1. Directive 2006/42/EC of the European Parliament and of the Council of 17 May 2006 on machinery, and amending Directive 95/16/EC (OJ L 157, 9.6.2006, p. 24);

+ 2. Directive 2009/48/EC of the European Parliament and of the Council of 18 June 2009 on the safety of toys (OJ L 170, 30.6.2009, p. 1);

+ 3. Directive 2013/53/EU of the European Parliament and of the Council of 20 November 2013 on recreational craft and personal watercraft and repealing Directive 94/25/EC (OJ L 354, 28.12.2013, p. 90);

+ 4. Directive 2014/33/EU of the European Parliament and of the Council of 26 February 2014 on the harmonisation of the laws of the Member States relating to lifts and safety components for lifts (OJ L 96, 29.3.2014, p. 251);

+ 5. Directive 2014/34/EU of the European Parliament and of the Council of 26 February 2014 on the harmonisation of the laws of the Member States relating to equipment and protective systems intended for use in potentially explosive atmospheres (OJ L 96, 29.3.2014, p. 309);

+ 6. Directive 2014/53/EU of the European Parliament and of the Council of 16 April 2014 on the harmonisation of the laws of the Member States relating to the making available on the market of radio equipment and repealing Directive 1999/5/EC (OJ L 153, 22.5.2014, p. 62);

+ 7. Directive 2014/68/EU of the European Parliament and of the Council of 15 May 2014 on the harmonisation of the laws of the Member States relating to the making available on the market of pressure equipment (OJ L 189, 27.6.2014, p. 164);

+ 8. Regulation (EU) 2016/424 of the European Parliament and of the Council of 9 March 2016 on cableway installations and repealing Directive 2000/9/EC (OJ L 81, 31.3.2016, p. 1);

+ 9. Regulation (EU) 2016/425 of the European Parliament and of the Council of 9 March 2016 on personal protective equipment and repealing Council Directive 89/686/EEC (OJ L 81, 31.3.2016, p. 51);

+ 10. Regulation (EU) 2016/426 of the European Parliament and of the Council of 9 March 2016 on appliances burning gaseous fuels and repealing Directive 2009/142/EC (OJ L 81, 31.3.2016, p. 99);

+ 11. Regulation (EU) 2017/745 of the European Parliament and of the Council of 5 April 2017 on medical devices, amending Directive 2001/83/EC, Regulation (EC) No 178/2002 and Regulation (EC) No 1223/2009 and repealing Council Directives 90/385/EEC and 93/42/EEC (OJ L 117, 5.5.2017, p. 1);

+ 12. Regulation (EU) 2017/746 of the European Parliament and of the Council of 5 April 2017 on in vitro diagnostic medical devices and repealing Directive 98/79/EC and Commission Decision 2010/227/EU (OJ L 117, 5.5.2017, p. 176).

+ Section B. List of other Union harmonisation legislation

+ 13. Regulation (EC) No 300/2008 of the European Parliament and of the Council of 11 March 2008 on common rules in the field of civil aviation security and repealing Regulation (EC) No 2320/2002 (OJ L 97, 9.4.2008, p. 72);

+ 14. Regulation (EU) No 168/2013 of the European Parliament and of the Council of 15 January 2013 on the approval and market surveillance of two- or three-wheel vehicles and quadricycles (OJ L 60, 2.3.2013, p. 52);

+ 15. Regulation (EU) No 167/2013 of the European Parliament and of the Council of 5 February 2013 on the approval and market surveillance of agricultural and forestry vehicles (OJ L 60, 2.3.2013, p. 1);

+ 124/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 16. Directive 2014/90/EU of the European Parliament and of the Council of 23 July 2014 on marine equipment and repealing Council Directive 96/98/EC (OJ L 257, 28.8.2014, p. 146);

+ 17. Directive (EU) 2016/797 of the European Parliament and of the Council of 11 May 2016 on the interoperability of the rail system within the European Union (OJ L 138, 26.5.2016, p. 44);

+ 18. Regulation (EU) 2018/858 of the European Parliament and of the Council of 30 May 2018 on the approval and market surveillance of motor vehicles and their trailers, and of systems, components and separate technical units intended for such vehicles, amending Regulations (EC) No 715/2007 and (EC) No 595/2009 and repealing Directive 2007/46/EC (OJ L 151, 14.6.2018, p. 1);

+ 19. Regulation (EU) 2019/2144 of the European Parliament and of the Council of 27 November 2019 on type-approval requirements for motor vehicles and their trailers, and systems, components and separate technical units intended for such vehicles, as regards their general safety and the protection of vehicle occupants and vulnerable road users, amending Regulation (EU) 2018/858 of the European Parliament and of the Council and repealing Regulations (EC) No 78/2009, (EC) No 79/2009 and (EC) No 661/2009 of the European Parliament and of the Council and Commission Regulations (EC) No 631/2009, (EU) No 406/2010, (EU) No 672/2010, (EU) No 1003/2010, (EU) No 1005/2010, (EU) No 1008/2010, (EU) No 1009/2010, (EU) No 19/2011, (EU) No 109/2011, (EU) No 458/2011, (EU) No 65/2012, (EU) No 130/2012, (EU) No 347/2012, (EU) No 351/2012, (EU) No 1230/2012 and (EU) 2015/166 (OJ L 325, 16.12.2019, p. 1);

+ 20. Regulation (EU) 2018/1139 of the European Parliament and of the Council of 4 July 2018 on common rules in the field of civil aviation and establishing a European Union Aviation Safety Agency, and amending Regulations (EC) No 2111/2005, (EC) No 1008/2008, (EU) No 996/2010, (EU) No 376/2014 and Directives 2014/30/EU and 2014/53/EU of the European Parliament and of the Council, and repealing Regulations (EC) No 552/2004 and (EC) No 216/2008 of the European Parliament and of the Council and Council Regulation (EEC) No 3922/91 (OJ L 212, 22.8.2018, p. 1), in so far as the design, production and placing on the market of aircrafts referred to in Article 2(1), points (a) and (b) thereof, where it concerns unmanned aircraft and their engines, propellers, parts and equipment to control them remotely, are concerned.

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 125/144

+ ANNEX II

+ List of criminal offences referred to in Article 5(1), first subparagraph, point (h)(iii)

+ Criminal offences referred to in Article 5(1), first subparagraph, point (h)(iii): -- terrorism, -- trafficking in human beings, -- sexual exploitation of children, and child pornography, -- illicit trafficking in narcotic drugs or psychotropic substances, -- illicit trafficking in weapons, munitions or explosives, -- murder, grievous bodily injury, -- illicit trade in human organs or tissue, -- illicit trafficking in nuclear or radioactive materials, -- kidnapping, illegal restraint or hostage-taking, -- crimes within the jurisdiction of the International Criminal Court, -- unlawful seizure of aircraft or ships, -- rape, -- environmental crime, -- organised or armed robbery, -- sabotage, -- participation in a criminal organisation involved in one or more of the offences listed above.

+ 126/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ ANNEX III

+ High-risk AI systems referred to in Article 6(2)

+ High-risk AI systems pursuant to Article 6(2) are the AI systems listed in any of the following areas:

+ 1. Biometrics, in so far as their use is permitted under relevant Union or national law:

+ (a) remote biometric identification systems.

+ This shall not include AI systems intended to be used for biometric verification the sole purpose of which is to confirm that a specific natural person is the person he or she claims to be;

+ (b) AI systems intended to be used for biometric categorisation, according to sensitive or protected attributes or characteristics based on the inference of those attributes or characteristics;

+ (c) AI systems intended to be used for emotion recognition.

+ 2. Critical infrastructure: AI systems intended to be used as safety components in the management and operation of critical digital infrastructure, road traffic, or in the supply of water, gas, heating or electricity.

+ 3. Education and vocational training:

+ (a) AI systems intended to be used to determine access or admission or to assign natural persons to educational and vocational training institutions at all levels;

+ (b) AI systems intended to be used to evaluate learning outcomes, including when those outcomes are used to steer the learning process of natural persons in educational and vocational training institutions at all levels;

+ (c) AI systems intended to be used for the purpose of assessing the appropriate level of education that an individual will receive or will be able to access, in the context of or within educational and vocational training institutions at all levels;

+ (d) AI systems intended to be used for monitoring and detecting prohibited behaviour of students during tests in the context of or within educational and vocational training institutions at all levels.

+ 4. Employment, workers' management and access to self-employment:

+ (a) AI systems intended to be used for the recruitment or selection of natural persons, in particular to place targeted job advertisements, to analyse and filter job applications, and to evaluate candidates;

+ (b) AI systems intended to be used to make decisions affecting terms of work-related relationships, the promotion or termination of work-related contractual relationships, to allocate tasks based on individual behaviour or personal traits or characteristics or to monitor and evaluate the performance and behaviour of persons in such relationships.

+ 5. Access to and enjoyment of essential private services and essential public services and benefits:

+ (a) AI systems intended to be used by public authorities or on behalf of public authorities to evaluate the eligibility of natural persons for essential public assistance benefits and services, including healthcare services, as well as to grant, reduce, revoke, or reclaim such benefits and services;

+ (b) AI systems intended to be used to evaluate the creditworthiness of natural persons or establish their credit score, with the exception of AI systems used for the purpose of detecting financial fraud;

+ (c) AI systems intended to be used for risk assessment and pricing in relation to natural persons in the case of life and health insurance;

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 127/144

+ (d) AI systems intended to evaluate and classify emergency calls by natural persons or to be used to dispatch, or to establish priority in the dispatching of, emergency first response services, including by police, firefighters and medical aid, as well as of emergency healthcare patient triage systems.

+ 6. Law enforcement, in so far as their use is permitted under relevant Union or national law:

+ (a) AI systems intended to be used by or on behalf of law enforcement authorities, or by Union institutions, bodies, offices or agencies in support of law enforcement authorities or on their behalf to assess the risk of a natural person becoming the victim of criminal offences;

+ (b) AI systems intended to be used by or on behalf of law enforcement authorities or by Union institutions, bodies, offices or agencies in support of law enforcement authorities as polygraphs or similar tools;

+ (c) AI systems intended to be used by or on behalf of law enforcement authorities, or by Union institutions, bodies, offices or agencies, in support of law enforcement authorities to evaluate the reliability of evidence in the course of the investigation or prosecution of criminal offences;

+ (d) AI systems intended to be used by law enforcement authorities or on their behalf or by Union institutions, bodies, offices or agencies in support of law enforcement authorities for assessing the risk of a natural person offending or re-offending not solely on the basis of the profiling of natural persons as referred to in Article 3(4) of Directive (EU) 2016/680, or to assess personality traits and characteristics or past criminal behaviour of natural persons or groups;

+ (e) AI systems intended to be used by or on behalf of law enforcement authorities or by Union institutions, bodies, offices or agencies in support of law enforcement authorities for the profiling of natural persons as referred to in Article 3(4) of Directive (EU) 2016/680 in the course of the detection, investigation or prosecution of criminal offences.

+ 7. Migration, asylum and border control management, in so far as their use is permitted under relevant Union or national law:

+ (a) AI systems intended to be used by or on behalf of competent public authorities or by Union institutions, bodies, offices or agencies as polygraphs or similar tools;

+ (b) AI systems intended to be used by or on behalf of competent public authorities or by Union institutions, bodies, offices or agencies to assess a risk, including a security risk, a risk of irregular migration, or a health risk, posed by a natural person who intends to enter or who has entered into the territory of a Member State;

+ (c) AI systems intended to be used by or on behalf of competent public authorities or by Union institutions, bodies, offices or agencies to assist competent public authorities for the examination of applications for asylum, visa or residence permits and for associated complaints with regard to the eligibility of the natural persons applying for a status, including related assessments of the reliability of evidence;

+ (d) AI systems intended to be used by or on behalf of competent public authorities, or by Union institutions, bodies, offices or agencies, in the context of migration, asylum or border control management, for the purpose of detecting, recognising or identifying natural persons, with the exception of the verification of travel documents.

+ 8. Administration of justice and democratic processes:

+ 128/144

+ (a) AI systems intended to be used by a judicial authority or on their behalf to assist a judicial authority in researching and interpreting facts and the law and in applying the law to a concrete set of facts, or to be used in a similar way in alternative dispute resolution;

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ (b) AI systems intended to be used for influencing the outcome of an election or referendum or the voting behaviour of natural persons in the exercise of their vote in elections or referenda. This does not include AI systems to the output of which natural persons are not directly exposed, such as tools used to organise, optimise or structure political campaigns from an administrative or logistical point of view.

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 129/144

+ ANNEX IV

+ Technical documentation referred to in Article 11(1)

+ The technical documentation referred to in Article 11(1) shall contain at least the following information, as applicable to the relevant AI system:

+ 1. A general description of the AI system including:

+ (a) its intended purpose, the name of the provider and the version of the system reflecting its relation to previous versions;

+ (b) how the AI system interacts with, or can be used to interact with, hardware or software, including with other AI systems, that are not part of the AI system itself, where applicable;

+ (c) the versions of relevant software or firmware, and any requirements related to version updates;

+ (d) the description of all the forms in which the AI system is placed on the market or put into service, such as software packages embedded into hardware, downloads, or APIs;

+ (e) the description of the hardware on which the AI system is intended to run;

+ (f) where the AI system is a component of products, photographs or illustrations showing external features, the marking and internal layout of those products;

+ (g) a basic description of the user-interface provided to the deployer;

+ (h) instructions for use for the deployer, and a basic description of the user-interface provided to the deployer, where applicable;

+ 2. A detailed description of the elements of the AI system and of the process for its development, including:

+ (a) the methods and steps performed for the development of the AI system, including, where relevant, recourse to pre-trained systems or tools provided by third parties and how those were used, integrated or modified by the provider;

+ (b) the design specifications of the system, namely the general logic of the AI system and of the algorithms; the key design choices including the rationale and assumptions made, including with regard to persons or groups of persons in respect of who, the system is intended to be used; the main classification choices; what the system is designed to optimise for, and the relevance of the different parameters; the description of the expected output and output quality of the system; the decisions about any possible trade-off made regarding the technical solutions adopted to comply with the requirements set out in Chapter III, Section 2;

+ (c) the description of the system architecture explaining how software components build on or feed into each other and integrate into the overall processing; the computational resources used to develop, train, test and validate the AI system;

+ (d) where relevant, the data requirements in terms of datasheets describing the training methodologies and techniques and the training data sets used, including a general description of these data sets, information about their provenance, scope and main characteristics; how the data was obtained and selected; labelling procedures (e.g. for supervised learning), data cleaning methodologies (e.g. outliers detection);

+ (e) assessment of the human oversight measures needed in accordance with Article 14, including an assessment of the technical measures needed to facilitate the interpretation of the outputs of AI systems by the deployers, in accordance with Article 13(3), point (d);

+ (f) where applicable, a detailed description of pre-determined changes to the AI system and its performance, together with all the relevant information related to the technical solutions adopted to ensure continuous compliance of the AI system with the relevant requirements set out in Chapter III, Section 2;

+ (g) the validation and testing procedures used, including information about the validation and testing data used and their main characteristics; metrics used to measure accuracy, robustness and compliance with other relevant requirements set out in Chapter III, Section 2, as well as potentially discriminatory impacts; test logs and all test reports dated and signed by the responsible persons, including with regard to pre-determined changes as referred to under point (f);

+ 130/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ (h) cybersecurity measures put in place;

+ 3. Detailed information about the monitoring, functioning and control of the AI system, in particular with regard to: its capabilities and limitations in performance, including the degrees of accuracy for specific persons or groups of persons on which the system is intended to be used and the overall expected level of accuracy in relation to its intended purpose; the foreseeable unintended outcomes and sources of risks to health and safety, fundamental rights and discrimination in view of the intended purpose of the AI system; the human oversight measures needed in accordance with Article 14, including the technical measures put in place to facilitate the interpretation of the outputs of AI systems by the deployers; specifications on input data, as appropriate;

+ 4. A description of the appropriateness of the performance metrics for the specific AI system;

+ 5. A detailed description of the risk management system in accordance with Article 9;

+ 6. A description of relevant changes made by the provider to the system through its lifecycle;

+ 7. A list of the harmonised standards applied in full or in part the references of which have been published in the Official Journal of the European Union; where no such harmonised standards have been applied, a detailed description of the solutions adopted to meet the requirements set out in Chapter III, Section 2, including a list of other relevant standards and technical specifications applied;

+ 8. A copy of the EU declaration of conformity referred to in Article 47;

+ 9. A detailed description of the system in place to evaluate the AI system performance in the post-market phase in accordance with Article 72, including the post-market monitoring plan referred to in Article 72(3).

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 131/144

+ ANNEX V

+ EU declaration of conformity

+ The EU declaration of conformity referred to in Article 47, shall contain all of the following information:

+ 1. AI system name and type and any additional unambiguous reference allowing the identification and traceability of the AI system;

+ 2. The name and address of the provider or, where applicable, of their authorised representative;

+ 3. A statement that the EU declaration of conformity referred to in Article 47 is issued under the sole responsibility of the provider;

+ 4. A statement that the AI system is in conformity with this Regulation and, if applicable, with any other relevant Union law that provides for the issuing of the EU declaration of conformity referred to in Article 47;

+ 5. Where an AI system involves the processing of personal data, a statement that that AI system complies with Regulations (EU) 2016/679 and (EU) 2018/1725 and Directive (EU) 2016/680;

+ 6. References to any relevant harmonised standards used or any other common specification in relation to which conformity is declared;

+ 7. Where applicable, the name and identification number of the notified body, a description of the conformity assessment procedure performed, and identification of the certificate issued;

+ 8. The place and date of issue of the declaration, the name and function of the person who signed it, as well as an indication for, or on behalf of whom, that person signed, a signature.

+ 132/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ ANNEX VI

+ Conformity assessment procedure based on internal control

+ 1. The conformity assessment procedure based on internal control is the conformity assessment procedure based on points 2, 3 and 4.

+ 2. The provider verifies that the established quality management system is in compliance with the requirements of Article 17.

+ 3. The provider examines the information contained in the technical documentation in order to assess the compliance of the AI system with the relevant essential requirements set out in Chapter III, Section 2.

+ 4. The provider also verifies that the design and development process of the AI system and its post-market monitoring as referred to in Article 72 is consistent with the technical documentation.

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 133/144

+ ANNEX VII

+ Conformity based on an assessment of the quality management system and an assessment of the technical documentation

+ 1. Introduction Conformity based on an assessment of the quality management system and an assessment of the technical documentation is the conformity assessment procedure based on points 2 to 5.

+ 2. Overview The approved quality management system for the design, development and testing of AI systems pursuant to Article 17 shall be examined in accordance with point 3 and shall be subject to surveillance as specified in point 5. The technical documentation of the AI system shall be examined in accordance with point 4.

+ 3. Quality management system 3.1. The application of the provider shall include:

+ (a) the name and address of the provider and, if the application is lodged by an authorised representative, also their name and address;

+ (b) the list of AI systems covered under the same quality management system;

+ (c) the technical documentation for each AI system covered under the same quality management system;

+ (d) the documentation concerning the quality management system which shall cover all the aspects listed under Article 17;

+ (e) a description of the procedures in place to ensure that the quality management system remains adequate and effective;

+ (f) a written declaration that the same application has not been lodged with any other notified body.

+ 3.2. The quality management system shall be assessed by the notified body, which shall determine whether it satisfies the requirements referred to in Article 17.

+ The decision shall be notified to the provider or its authorised representative.

+ The notification shall contain the conclusions of the assessment of the quality management system and the reasoned assessment decision.

+ 3.3. The quality management system as approved shall continue to be implemented and maintained by the provider so that it remains adequate and efficient.

+ 3.4. Any intended change to the approved quality management system or the list of AI systems covered by the latter shall be brought to the attention of the notified body by the provider.

+ The proposed changes shall be examined by the notified body, which shall decide whether the modified quality management system continues to satisfy the requirements referred to in point 3.2 or whether a reassessment is necessary.

+ The notified body shall notify the provider of its decision. The notification shall contain the conclusions of the examination of the changes and the reasoned assessment decision.

+ 4. Control of the technical documentation. 4.1. In addition to the application referred to in point 3, an application with a notified body of their choice shall be

+ lodged by the provider for the assessment of the technical documentation relating to the AI system which the provider intends to place on the market or put into service and which is covered by the quality management system referred to under point 3.

+ 4.2. The application shall include:

+ (a) the name and address of the provider;

+ (b) a written declaration that the same application has not been lodged with any other notified body;

+ (c) the technical documentation referred to in Annex IV.

+ 134/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 4.3. The technical documentation shall be examined by the notified body. Where relevant, and limited to what is necessary to fulfil its tasks, the notified body shall be granted full access to the training, validation, and testing data sets used, including, where appropriate and subject to security safeguards, through API or other relevant technical means and tools enabling remote access.

+ 4.4. In examining the technical documentation, the notified body may require that the provider supply further evidence or carry out further tests so as to enable a proper assessment of the conformity of the AI system with the requirements set out in Chapter III, Section 2. Where the notified body is not satisfied with the tests carried out by the provider, the notified body shall itself directly carry out adequate tests, as appropriate.

+ 4.5. Where necessary to assess the conformity of the high-risk AI system with the requirements set out in Chapter III, Section 2, after all other reasonable means to verify conformity have been exhausted and have proven to be insufficient, and upon a reasoned request, the notified body shall also be granted access to the training and trained models of the AI system, including its relevant parameters. Such access shall be subject to existing Union law on the protection of intellectual property and trade secrets.

+ 4.6. The decision of the notified body shall be notified to the provider or its authorised representative. The notification shall contain the conclusions of the assessment of the technical documentation and the reasoned assessment decision.

+ Where the AI system is in conformity with the requirements set out in Chapter III, Section 2, the notified body shall issue a Union technical documentation assessment certificate. The certificate shall indicate the name and address of the provider, the conclusions of the examination, the conditions (if any) for its validity and the data necessary for the identification of the AI system.

+ The certificate and its annexes shall contain all relevant information to allow the conformity of the AI system to be evaluated, and to allow for control of the AI system while in use, where applicable.

+ Where the AI system is not in conformity with the requirements set out in Chapter III, Section 2, the notified body shall refuse to issue a Union technical documentation assessment certificate and shall inform the applicant accordingly, giving detailed reasons for its refusal.

+ Where the AI system does not meet the requirement relating to the data used to train it, re-training of the AI system will be needed prior to the application for a new conformity assessment. In this case, the reasoned assessment decision of the notified body refusing to issue the Union technical documentation assessment certificate shall contain specific considerations on the quality data used to train the AI system, in particular on the reasons for non-compliance.

+ 4.7. Any change to the AI system that could affect the compliance of the AI system with the requirements or its intended purpose shall be assessed by the notified body which issued the Union technical documentation assessment certificate. The provider shall inform such notified body of its intention to introduce any of the abovementioned changes, or if it otherwise becomes aware of the occurrence of such changes. The intended changes shall be assessed by the notified body, which shall decide whether those changes require a new conformity assessment in accordance with Article 43(4) or whether they could be addressed by means of a supplement to the Union technical documentation assessment certificate. In the latter case, the notified body shall assess the changes, notify the provider of its decision and, where the changes are approved, issue to the provider a supplement to the Union technical documentation assessment certificate.

+ 5. Surveillance of the approved quality management system.

+ 5.1. The purpose of the surveillance carried out by the notified body referred to in Point 3 is to make sure that the provider duly complies with the terms and conditions of the approved quality management system.

+ 5.2. For assessment purposes, the provider shall allow the notified body to access the premises where the design, development, testing of the AI systems is taking place. The provider shall further share with the notified body all necessary information.

+ 5.3. The notified body shall carry out periodic audits to make sure that the provider maintains and applies the quality management system and shall provide the provider with an audit report. In the context of those audits, the notified body may carry out additional tests of the AI systems for which a Union technical documentation assessment certificate was issued.

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 135/144

+ ANNEX VIII

+ Information to be submitted upon the registration of high-risk AI systems in accordance with Article 49

+ Section A -- Information to be submitted by providers of high-risk AI systems in accordance with Article 49(1) The following information shall be provided and thereafter kept up to date with regard to high-risk AI systems to be registered in accordance with Article 49(1):

+ 1. The name, address and contact details of the provider;

+ 2. Where submission of information is carried out by another person on behalf of the provider, the name, address and contact details of that person;

+ 3. The name, address and contact details of the authorised representative, where applicable;

+ 4. The AI system trade name and any additional unambiguous reference allowing the identification and traceability of the AI system;

+ 5. A description of the intended purpose of the AI system and of the components and functions supported through this AI system;

+ 6. A basic and concise description of the information used by the system (data, inputs) and its operating logic;

+ 7. The status of the AI system (on the market, or in service; no longer placed on the market/in service, recalled);

+ 8. The type, number and expiry date of the certificate issued by the notified body and the name or identification number of that notified body, where applicable;

+ 9. A scanned copy of the certificate referred to in point 8, where applicable;

+ 10. Any Member States in which the AI system has been placed on the market, put into service or made available in the Union;

+ 11. A copy of the EU declaration of conformity referred to in Article 47;

+ 12. Electronic instructions for use; this information shall not be provided for high-risk AI systems in the areas of law enforcement or migration, asylum and border control management referred to in Annex III, points 1, 6 and 7;

+ 13. A URL for additional information (optional).

+ Section B -- Information to be submitted by providers of high-risk AI systems in accordance with Article 49(2) The following information shall be provided and thereafter kept up to date with regard to AI systems to be registered in accordance with Article 49(2):

+ 1. The name, address and contact details of the provider;

+ 2. Where submission of information is carried out by another person on behalf of the provider, the name, address and contact details of that person;

+ 3. The name, address and contact details of the authorised representative, where applicable;

+ 4. The AI system trade name and any additional unambiguous reference allowing the identification and traceability of the AI system;

+ 5. A description of the intended purpose of the AI system;

+ 6. The condition or conditions under Article 6(3)based on which the AI system is considered to be not-high-risk;

+ 7. A short summary of the grounds on which the AI system is considered to be not-high-risk in application of the procedure under Article 6(3);

+ 8. The status of the AI system (on the market, or in service; no longer placed on the market/in service, recalled);

+ 9. Any Member States in which the AI system has been placed on the market, put into service or made available in the Union.

+ 136/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ Section C -- Information to be submitted by deployers of high-risk AI systems in accordance with Article 49(3)

+ The following information shall be provided and thereafter kept up to date with regard to high-risk AI systems to be registered in accordance with Article 49(3):

+ 1. The name, address and contact details of the deployer;

+ 2. The name, address and contact details of the person submitting information on behalf of the deployer;

+ 3. The URL of the entry of the AI system in the EU database by its provider;

+ 4. A summary of the findings of the fundamental rights impact assessment conducted in accordance with Article 27;

+ 5. A summary of the data protection impact assessment carried out in accordance with Article 35 of Regulation (EU) 2016/679 or Article 27 of Directive (EU) 2016/680 as specified in Article 26(8) of this Regulation, where applicable.

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 137/144

+ ANNEX IX

+ Information to be submitted upon the registration of high-risk AI systems listed in Annex III in relation to testing in real world conditions in accordance with Article 60

+ The following information shall be provided and thereafter kept up to date with regard to testing in real world conditions to be registered in accordance with Article 60:

+ 1. A Union-wide unique single identification number of the testing in real world conditions;

+ 2. The name and contact details of the provider or prospective provider and of the deployers involved in the testing in real world conditions;

+ 3. A brief description of the AI system, its intended purpose, and other information necessary for the identification of the system;

+ 4. A summary of the main characteristics of the plan for testing in real world conditions;

+ 5. Information on the suspension or termination of the testing in real world conditions.

+ 138/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ OJ L, 12.7.2024

+ ANNEX X Union legislative acts on large-scale IT systems in the area of Freedom, Security and Justice

+ EN

+ 1. Schengen Information System

+ (a) Regulation (EU) 2018/1860 of the European Parliament and of the Council of 28 November 2018 on the use of the Schengen Information System for the return of illegally staying third-country nationals (OJ L 312, 7.12.2018, p. 1).

+ (b) Regulation (EU) 2018/1861 of the European Parliament and of the Council of 28 November 2018 on the establishment, operation and use of the Schengen Information System (SIS) in the field of border checks, and amending the Convention implementing the Schengen Agreement, and amending and repealing Regulation (EC) No 1987/2006 (OJ L 312, 7.12.2018, p. 14).

+ (c) Regulation (EU) 2018/1862 of the European Parliament and of the Council of 28 November 2018 on the establishment, operation and use of the Schengen Information System (SIS) in the field of police cooperation and judicial cooperation in criminal matters, amending and repealing Council Decision 2007/533/JHA, and repealing Regulation (EC) No 1986/2006 of the European Parliament and of the Council and Commission Decision 2010/261/EU (OJ L 312, 7.12.2018, p. 56).

+ 2. Visa Information System

+ (a) Regulation (EU) 2021/1133 of the European Parliament and of the Council of 7 July 2021 amending Regulations (EU) No 603/2013, (EU) 2016/794, (EU) 2018/1862, (EU) 2019/816 and (EU) 2019/818 as regards the establishment of the conditions for accessing other EU information systems for the purposes of the Visa Information System (OJ L 248, 13.7.2021, p. 1).

+ (b) Regulation (EU) 2021/1134 of the European Parliament and of the Council of 7 July 2021 amending Regulations (EC) No 767/2008, (EC) No 810/2009, (EU) 2016/399, (EU) 2017/2226, (EU) 2018/1240, (EU) 2018/1860, (EU) 2018/1861, (EU) 2019/817 and (EU) 2019/1896 of the European Parliament and of the Council and repealing Council Decisions 2004/512/EC and 2008/633/JHA, for the purpose of reforming the Visa Information System (OJ L 248, 13.7.2021, p. 11).

+ 3. Eurodac

+ Regulation (EU) 2024/1358 of the European Parliament and of the Council of 14 May 2024 on the establishment of `Eurodac' for the comparison of biometric data in order to effectively apply Regulations (EU) 2024/1315 and (EU) 2024/1350 of the European Parliament and of the Council and Council Directive 2001/55/EC and to identify illegally staying third-country nationals and stateless persons and on requests for the comparison with Eurodac data by Member States' law enforcement authorities and Europol for law enforcement purposes, amending Regulations (EU) 2018/1240 and (EU) 2019/818 of the European Parliament and of the Council and repealing Regulation (EU) No 603/2013 of the European Parliament and of the Council (OJ L, 2024/1358, 22.5.2024, ELI: http://data.europa. eu/eli/reg/2024/1358/oj).

+ 4. Entry/Exit System

+ Regulation (EU) 2017/2226 of the European Parliament and of the Council of 30 November 2017 establishing an Entry/Exit System (EES) to register entry and exit data and refusal of entry data of third-country nationals crossing the external borders of the Member States and determining the conditions for access to the EES for law enforcement purposes, and amending the Convention implementing the Schengen Agreement and Regulations (EC) No 767/2008 and (EU) No 1077/2011 (OJ L 327, 9.12.2017, p. 20).

+ 5. European Travel Information and Authorisation System

+ (a) Regulation (EU) 2018/1240 of the European Parliament and of the Council of 12 September 2018 establishing a European Travel Information and Authorisation System (ETIAS) and amending Regulations (EU) No 1077/2011, (EU) No 515/2014, (EU) 2016/399, (EU) 2016/1624 and (EU) 2017/2226 (OJ L 236, 19.9.2018, p. 1).

+ (b) Regulation (EU) 2018/1241 of the European Parliament and of the Council of 12 September 2018 amending Regulation (EU) 2016/794 for the purpose of establishing a European Travel Information and Authorisation System (ETIAS) (OJ L 236, 19.9.2018, p. 72).

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 139/144

+ EN

+ OJ L, 12.7.2024

+ 6. European Criminal Records Information System on third-country nationals and stateless persons

+ Regulation (EU) 2019/816 of the European Parliament and of the Council of 17 April 2019 establishing a centralised system for the identification of Member States holding conviction information on third-country nationals and stateless persons (ECRIS-TCN) to supplement the European Criminal Records Information System and amending Regulation (EU) 2018/1726 (OJ L 135, 22.5.2019, p. 1).

+ 7. Interoperability

+ (a) Regulation (EU) 2019/817 of the European Parliament and of the Council of 20 May 2019 on establishing a framework for interoperability between EU information systems in the field of borders and visa and amending Regulations (EC) No 767/2008, (EU) 2016/399, (EU) 2017/2226, (EU) 2018/1240, (EU) 2018/1726 and (EU) 2018/1861 of the European Parliament and of the Council and Council Decisions 2004/512/EC and 2008/633/JHA (OJ L 135, 22.5.2019, p. 27).

+ (b) Regulation (EU) 2019/818 of the European Parliament and of the Council of 20 May 2019 on establishing a framework for interoperability between EU information systems in the field of police and judicial cooperation, asylum and migration and amending Regulations (EU) 2018/1726, (EU) 2018/1862 and (EU) 2019/816 (OJ L 135, 22.5.2019, p. 85).

+ 140/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ ANNEX XI

+ Technical documentation referred to in Article 53(1), point (a) -- technical documentation for providers of general-purpose AI models

+ Section 1

+ Information to be provided by all providers of general-purpose AI models

+ The technical documentation referred to in Article 53(1), point (a) shall contain at least the following information as appropriate to the size and risk profile of the model:

+ 1. A general description of the general-purpose AI model including:

+ (a) the tasks that the model is intended to perform and the type and nature of AI systems in which it can be integrated;

+ (b) the acceptable use policies applicable;

+ (c) the date of release and methods of distribution;

+ (d) the architecture and number of parameters;

+ (e) the modality (e.g. text, image) and format of inputs and outputs;

+ (f) the licence.

+ 2. A detailed description of the elements of the model referred to in point 1, and relevant information of the process for the development, including the following elements:

+ (a) the technical means (e.g. instructions of use, infrastructure, tools) required for the general-purpose AI model to be integrated in AI systems;

+ (b) the design specifications of the model and training process, including training methodologies and techniques, the key design choices including the rationale and assumptions made; what the model is designed to optimise for and the relevance of the different parameters, as applicable;

+ (c) information on the data used for training, testing and validation, where applicable, including the type and provenance of data and curation methodologies (e.g. cleaning, filtering, etc.), the number of data points, their scope and main characteristics; how the data was obtained and selected as well as all other measures to detect the unsuitability of data sources and methods to detect identifiable biases, where applicable;

+ (d) the computational resources used to train the model (e.g. number of floating point operations), training time, and other relevant details related to the training;

+ (e) known or estimated energy consumption of the model.

+ With regard to point (e), where the energy consumption of the model is unknown, the energy consumption may be based on information about computational resources used.

+ Section 2

+ Additional information to be provided by providers of general-purpose AI models with systemic risk

+ 1. A detailed description of the evaluation strategies, including evaluation results, on the basis of available public evaluation protocols and tools or otherwise of other evaluation methodologies. Evaluation strategies shall include evaluation criteria, metrics and the methodology on the identification of limitations.

+ 2. Where applicable, a detailed description of the measures put in place for the purpose of conducting internal and/or external adversarial testing (e.g. red teaming), model adaptations, including alignment and fine-tuning.

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 141/144

+ 3. Where applicable, a detailed description of the system architecture explaining how software components build or feed into each other and integrate into the overall processing.

+ 142/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ ANNEX XII

+ Transparency information referred to in Article 53(1), point (b) -- technical documentation for providers of general-purpose AI models to downstream providers that integrate the model into their

+ AI system

+ The information referred to in Article 53(1), point (b) shall contain at least the following:

+ 1. A general description of the general-purpose AI model including:

+ (a) the tasks that the model is intended to perform and the type and nature of AI systems into which it can be integrated;

+ (b) the acceptable use policies applicable;

+ (c) the date of release and methods of distribution;

+ (d) how the model interacts, or can be used to interact, with hardware or software that is not part of the model itself, where applicable;

+ (e) the versions of relevant software related to the use of the general-purpose AI model, where applicable;

+ (f) the architecture and number of parameters;

+ (g) the modality (e.g. text, image) and format of inputs and outputs;

+ (h) the licence for the model.

+ 2. A description of the elements of the model and of the process for its development, including:

+ (a) the technical means (e.g. instructions for use, infrastructure, tools) required for the general-purpose AI model to be integrated into AI systems;

+ (b) the modality (e.g. text, image, etc.) and format of the inputs and outputs and their maximum size (e.g. context window length, etc.);

+ (c) information on the data used for training, testing and validation, where applicable, including the type and provenance of data and curation methodologies.

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

+ 143/144

+ Criteria for the designation of general-purpose AI models with systemic risk referred to in Article 51

+ For the purpose of determining that a general-purpose AI model has capabilities or an impact equivalent to those set out in Article 51(1), point (a), the Commission shall take into account the following criteria:

+ (a) the number of parameters of the model;

+ (b) the quality or size of the data set, for example measured through tokens;

+ (c) the amount of computation used for training the model, measured in floating point operations or indicated by a combination of other variables such as estimated cost of training, estimated time required for the training, or estimated energy consumption for the training;

+ (d) the input and output modalities of the model, such as text to text (large language models), text to image, multi-modality, and the state of the art thresholds for determining high-impact capabilities for each modality, and the specific type of inputs and outputs (e.g. biological sequences);

+ (e) the benchmarks and evaluations of capabilities of the model, including considering the number of tasks without additional training, adaptability to learn new, distinct tasks, its level of autonomy and scalability, the tools it has access to;

+ (f) whether it has a high impact on the internal market due to its reach, which shall be presumed when it has been made available to at least 10 000 registered business users established in the Union;

+ (g) the number of registered end-users.

+ 144/144

+ ELI: http://data.europa.eu/eli/reg/2024/1689/oj

This is a computed comparison between saved source copies. It is not a statement by the source, a legal conclusion, or a determination of current status.