Recitals

The recitals set out the reasoning behind the Act's articles.

Recital 1 — (1) The purpose of this Regulation is to improve the functioning of the internal market by laying down a unifo…
Recital 2 — (2) This Regulation should be applied in accordance with the values of the Union enshrined as in the Charter, …
Recital 3 — (3) AI systems can be easily deployed in a large variety of sectors of the economy and many parts of society, …
Recital 4 — (4) AI is a fast evolving family of technologies that contributes to a wide array of economic, environmental a…
Recital 5 — (5) At the same time, depending on the circumstances regarding its specific application, use, and level of tec…
Recital 6 — (6) Given the major impact that AI can have on society and the need to build trust, it is vital for AI and its…
Recital 7 — (7) In order to ensure a consistent and high level of protection of public interests as regards health, safety…
Recital 8 — (8) A Union legal framework laying down harmonised rules on AI is therefore needed to foster the development, …
Recital 9 — (9) Harmonised rules applicable to the placing on the market, the putting into service and the use of high-ris…
Recital 10 — (10) The fundamental right to the protection of personal data is safeguarded in particular by Regulations (EU)…
Recital 11 — (11) This Regulation should be without prejudice to the provisions regarding the liability of providers of int…
Recital 12 — (12) The notion of ‘AI system’ in this Regulation should be clearly defined and should be closely aligned with…
Recital 13 — (13) The notion of ‘deployer’ referred to in this Regulation should be interpreted as any natural or legal per…
Recital 14 — (14) The notion of ‘biometric data’ used in this Regulation should be interpreted in light of the notion of bi…
Recital 15 — (15) The notion of ‘biometric identification’ referred to in this Regulation should be defined as the automate…
Recital 16 — (16) The notion of ‘biometric categorisation’ referred to in this Regulation should be defined as assigning na…
Recital 17 — (17) The notion of ‘remote biometric identification system’ referred to in this Regulation should be defined f…
Recital 18 — (18) The notion of ‘emotion recognition system’ referred to in this Regulation should be defined as an AI syst…
Recital 19 — (19) For the purposes of this Regulation the notion of ‘publicly accessible space’ should be understood as ref…
Recital 20 — (20) In order to obtain the greatest benefits from AI systems while protecting fundamental rights, health and …
Recital 21 — (21) In order to ensure a level playing field and an effective protection of rights and freedoms of individual…
Recital 22 — (22) In light of their digital nature, certain AI systems should fall within the scope of this Regulation even…
Recital 23 — (23) This Regulation should also apply to Union institutions, bodies, offices and agencies when acting as a pr…
Recital 24 — (24) If, and insofar as, AI systems are placed on the market, put into service, or used with or without modifi…
Recital 25 — (25) This Regulation should support innovation, should respect freedom of science, and should not undermine re…
Recital 26 — (26) In order to introduce a proportionate and effective set of binding rules for AI systems, a clearly define…
Recital 27 — (27) While the risk-based approach is the basis for a proportionate and effective set of binding rules, it is …
Recital 28 — (28) Aside from the many beneficial uses of AI, it can also be misused and provide novel and powerful tools fo…
Recital 29 — (29) AI-enabled manipulative techniques can be used to persuade persons to engage in unwanted behaviours, or t…
Recital 30 — (30) Biometric categorisation systems that are based on natural persons’ biometric data, such as an individual…
Recital 31 — (31) AI systems providing social scoring of natural persons by public or private actors may lead to discrimina…
Recital 32 — (32) The use of AI systems for ‘real-time’ remote biometric identification of natural persons in publicly acce…
Recital 33 — (33) The use of those systems for the purpose of law enforcement should therefore be prohibited, except in exh…
Recital 34 — (34) In order to ensure that those systems are used in a responsible and proportionate manner, it is also impo…
Recital 35 — (35) Each use of a ‘real-time’ remote biometric identification system in publicly accessible spaces for the pu…
Recital 36 — (36) In order to carry out their tasks in accordance with the requirements set out in this Regulation as well …
Recital 37 — (37) Furthermore, it is appropriate to provide, within the exhaustive framework set by this Regulation that su…
Recital 38 — (38) The use of AI systems for real-time remote biometric identification of natural persons in publicly access…
Recital 39 — (39) Any processing of biometric data and other personal data involved in the use of AI systems for biometric …
Recital 40 — (40) In accordance with Article 6a of Protocol No 21 on the position of the United Kingdom and Ireland in resp…
Recital 41 — (41) In accordance with Articles 2 and 2a of Protocol No 22 on the position of Denmark, annexed to the TEU and…
Recital 42 — (42) In line with the presumption of innocence, natural persons in the Union should always be judged on their …
Recital 43 — (43) The placing on the market, the putting into service for that specific purpose, or the use of AI systems t…
Recital 44 — (44) There are serious concerns about the scientific basis of AI systems aiming to identify or infer emotions,…
Recital 45 — (45) Practices that are prohibited by Union law, including data protection law, non-discrimination law, consum…
Recital 46 — (46) High-risk AI systems should only be placed on the Union market, put into service or used if they comply w…
Recital 47 — (47) AI systems could have an adverse impact on the health and safety of persons, in particular when such syst…
Recital 48 — (48) The extent of the adverse impact caused by the AI system on the fundamental rights protected by the Chart…
Recital 49 — (49) As regards high-risk AI systems that are safety components of products or systems, or which are themselve…
Recital 50 — (50) As regards AI systems that are safety components of products, or which are themselves products, falling w…
Recital 51 — (51) The classification of an AI system as high-risk pursuant to this Regulation should not necessarily mean t…
Recital 52 — (52) As regards stand-alone AI systems, namely high-risk AI systems other than those that are safety component…
Recital 53 — (53) It is also important to clarify that there may be specific cases in which AI systems referred to in pre-d…
Recital 54 — (54) As biometric data constitutes a special category of personal data, it is appropriate to classify as high-…
Recital 55 — (55) As regards the management and operation of critical infrastructure, it is appropriate to classify as high…
Recital 56 — (56) The deployment of AI systems in education is important to promote high-quality digital education and trai…
Recital 57 — (57) AI systems used in employment, workers management and access to self-employment, in particular for the re…
Recital 58 — (58) Another area in which the use of AI systems deserves special consideration is the access to and enjoyment…
Recital 59 — (59) Given their role and responsibility, actions by law enforcement authorities involving certain uses of AI …
Recital 60 — (60) AI systems used in migration, asylum and border control management affect persons who are often in partic…
Recital 61 — (61) Certain AI systems intended for the administration of justice and democratic processes should be classifi…
Recital 62 — (62) Without prejudice to the rules provided for in Regulation (EU) 2024/900 of the European Parliament and of…
Recital 63 — (63) The fact that an AI system is classified as a high-risk AI system under this Regulation should not be int…
Recital 64 — (64) To mitigate the risks from high-risk AI systems placed on the market or put into service and to ensure a …
Recital 65 — (65) The risk-management system should consist of a continuous, iterative process that is planned and run thro…
Recital 66 — (66) Requirements should apply to high-risk AI systems as regards risk management, the quality and relevance o…
Recital 67 — (67) High-quality data and access to high-quality data plays a vital role in providing structure and in ensuri…
Recital 68 — (68) For the development and assessment of high-risk AI systems, certain actors, such as providers, notified b…
Recital 69 — (69) The right to privacy and to protection of personal data must be guaranteed throughout the entire lifecycl…
Recital 70 — (70) In order to protect the right of others from the discrimination that might result from the bias in AI sys…
Recital 71 — (71) Having comprehensible information on how high-risk AI systems have been developed and how they perform th…
Recital 72 — (72) To address concerns related to opacity and complexity of certain AI systems and help deployers to fulfil …
Recital 73 — (73) High-risk AI systems should be designed and developed in such a way that natural persons can oversee thei…
Recital 74 — (74) High-risk AI systems should perform consistently throughout their lifecycle and meet an appropriate level…
Recital 75 — (75) Technical robustness is a key requirement for high-risk AI systems. They should be resilient in relation …
Recital 76 — (76) Cybersecurity plays a crucial role in ensuring that AI systems are resilient against attempts to alter th…
Recital 77 — (77) Without prejudice to the requirements related to robustness and accuracy set out in this Regulation, high…
Recital 78 — (78) The conformity assessment procedure provided by this Regulation should apply in relation to the essential…
Recital 79 — (79) It is appropriate that a specific natural or legal person, defined as the provider, takes responsibility …
Recital 80 — (80) As signatories to the United Nations Convention on the Rights of Persons with Disabilities, the Union and…
Recital 81 — (81) The provider should establish a sound quality management system, ensure the accomplishment of the require…
Recital 82 — (82) To enable enforcement of this Regulation and create a level playing field for operators, and, taking into…
Recital 83 — (83) In light of the nature and complexity of the value chain for AI systems and in line with the New Legislat…
Recital 84 — (84) To ensure legal certainty, it is necessary to clarify that, under certain specific conditions, any distri…
Recital 85 — (85) General-purpose AI systems may be used as high-risk AI systems by themselves or be components of other hi…
Recital 86 — (86) Where, under the conditions laid down in this Regulation, the provider that initially placed the AI syste…
Recital 87 — (87) In addition, where a high-risk AI system that is a safety component of a product which falls within the s…
Recital 88 — (88) Along the AI value chain multiple parties often supply AI systems, tools and services but also components…
Recital 89 — (89) Third parties making accessible to the public tools, services, processes, or AI components other than gen…
Recital 90 — (90) The Commission could develop and recommend voluntary model contractual terms between providers of high-ri…
Recital 91 — (91) Given the nature of AI systems and the risks to safety and fundamental rights possibly associated with th…
Recital 92 — (92) This Regulation is without prejudice to obligations for employers to inform or to inform and consult work…
Recital 93 — (93) Whilst risks related to AI systems can result from the way such systems are designed, risks can as well s…
Recital 94 — (94) Any processing of biometric data involved in the use of AI systems for biometric identification for the p…
Recital 95 — (95) Without prejudice to applicable Union law, in particular Regulation (EU) 2016/679 and Directive (EU) 2016…
Recital 96 — (96) In order to efficiently ensure that fundamental rights are protected, deployers of high-risk AI systems t…
Recital 97 — (97) The notion of general-purpose AI models should be clearly defined and set apart from the notion of AI sys…
Recital 98 — (98) Whereas the generality of a model could, inter alia, also be determined by a number of parameters, models…
Recital 99 — (99) Large generative AI models are a typical example for a general-purpose AI model, given that they allow fo…
Recital 100 — (100) When a general-purpose AI model is integrated into or forms part of an AI system, this system should be …
Recital 101 — (101) Providers of general-purpose AI models have a particular role and responsibility along the AI value chai…
Recital 102 — (102) Software and data, including models, released under a free and open-source licence that allows them to b…
Recital 103 — (103) Free and open-source AI components covers the software and data, including models and general-purpose AI…
Recital 104 — (104) The providers of general-purpose AI models that are released under a free and open-source licence, and w…
Recital 105 — (105) General-purpose AI models, in particular large generative AI models, capable of generating text, images,…
Recital 106 — (106) Providers that place general-purpose AI models on the Union market should ensure compliance with the rel…
Recital 107 — (107) In order to increase transparency on the data that is used in the pre-training and training of general-p…
Recital 108 — (108) With regard to the obligations imposed on providers of general-purpose AI models to put in place a polic…
Recital 109 — (109) Compliance with the obligations applicable to the providers of general-purpose AI models should be comme…
Recital 110 — (110) General-purpose AI models could pose systemic risks which include, but are not limited to, any actual or…
Recital 111 — (111) It is appropriate to establish a methodology for the classification of general-purpose AI models as gene…
Recital 112 — (112) It is also necessary to clarify a procedure for the classification of a general-purpose AI model with sy…
Recital 113 — (113) If the Commission becomes aware of the fact that a general-purpose AI model meets the requirements to cl…
Recital 114 — (114) The providers of general-purpose AI models presenting systemic risks should be subject, in addition to t…
Recital 115 — (115) Providers of general-purpose AI models with systemic risks should assess and mitigate possible systemic …
Recital 116 — (116) The AI Office should encourage and facilitate the drawing up, review and adaptation of codes of practice…
Recital 117 — (117) The codes of practice should represent a central tool for the proper compliance with the obligations pro…
Recital 118 — (118) This Regulation regulates AI systems and AI models by imposing certain requirements and obligations for …
Recital 119 — (119) Considering the quick pace of innovation and the technological evolution of digital services in scope of…
Recital 120 — (120) Furthermore, obligations placed on providers and deployers of certain AI systems in this Regulation to e…
Recital 121 — (121) Standardisation should play a key role to provide technical solutions to providers to ensure compliance …
Recital 122 — (122) It is appropriate that, without prejudice to the use of harmonised standards and common specifications, …
Recital 123 — (123) In order to ensure a high level of trustworthiness of high-risk AI systems, those systems should be subj…
Recital 124 — (124) It is appropriate that, in order to minimise the burden on operators and avoid any possible duplication,…
Recital 125 — (125) Given the complexity of high-risk AI systems and the risks that are associated with them, it is importan…
Recital 126 — (126) In order to carry out third-party conformity assessments when so required, notified bodies should be not…
Recital 127 — (127) In line with Union commitments under the World Trade Organization Agreement on Technical Barriers to Tra…
Recital 128 — (128) In line with the commonly established notion of substantial modification for products regulated by Union…
Recital 129 — (129) High-risk AI systems should bear the CE marking to indicate their conformity with this Regulation so tha…
Recital 130 — (130) Under certain conditions, rapid availability of innovative technologies may be crucial for health and sa…
Recital 131 — (131) In order to facilitate the work of the Commission and the Member States in the AI field as well as to in…
Recital 132 — (132) Certain AI systems intended to interact with natural persons or to generate content may pose specific ri…
Recital 133 — (133) A variety of AI systems can generate large quantities of synthetic content that becomes increasingly har…
Recital 134 — (134) Further to the technical solutions employed by the providers of the AI system, deployers who use an AI s…
Recital 135 — (135) Without prejudice to the mandatory nature and full applicability of the transparency obligations, the Co…
Recital 136 — (136) The obligations placed on providers and deployers of certain AI systems in this Regulation to enable the…
Recital 137 — (137) Compliance with the transparency obligations for the AI systems covered by this Regulation should not be…
Recital 138 — (138) AI is a rapidly developing family of technologies that requires regulatory oversight and a safe and cont…
Recital 139 — (139) The objectives of the AI regulatory sandboxes should be to foster AI innovation by establishing a contro…
Recital 140 — (140) This Regulation should provide the legal basis for the providers and prospective providers in the AI reg…
Recital 141 — (141) In order to accelerate the process of development and the placing on the market of the high-risk AI syst…
Recital 142 — (142) To ensure that AI leads to socially and environmentally beneficial outcomes, Member States are encourage…
Recital 143 — (143) In order to promote and protect innovation, it is important that the interests of SMEs, including start-…
Recital 144 — (144) In order to promote and protect innovation, the AI-on-demand platform, all relevant Union funding progra…
Recital 145 — (145) In order to minimise the risks to implementation resulting from lack of knowledge and expertise in the m…
Recital 146 — (146) Moreover, in light of the very small size of some operators and in order to ensure proportionality regar…
Recital 147 — (147) It is appropriate that the Commission facilitates, to the extent possible, access to testing and experim…
Recital 148 — (148) This Regulation should establish a governance framework that both allows to coordinate and support the a…
Recital 149 — (149) In order to facilitate a smooth, effective and harmonised implementation of this Regulation a Board shou…
Recital 150 — (150) With a view to ensuring the involvement of stakeholders in the implementation and application of this Re…
Recital 151 — (151) To support the implementation and enforcement of this Regulation, in particular the monitoring activitie…
Recital 152 — (152) In order to support adequate enforcement as regards AI systems and reinforce the capacities of the Membe…
Recital 153 — (153) Member States hold a key role in the application and enforcement of this Regulation. In that respect, ea…
Recital 154 — (154) The national competent authorities should exercise their powers independently, impartially and without b…
Recital 155 — (155) In order to ensure that providers of high-risk AI systems can take into account the experience on the us…
Recital 156 — (156) In order to ensure an appropriate and effective enforcement of the requirements and obligations set out …
Recital 157 — (157) This Regulation is without prejudice to the competences, tasks, powers and independence of relevant nati…
Recital 158 — (158) Union financial services law includes internal governance and risk-management rules and requirements whi…
Recital 159 — (159) Each market surveillance authority for high-risk AI systems in the area of biometrics, as listed in an a…
Recital 160 — (160) The market surveillance authorities and the Commission should be able to propose joint activities, inclu…
Recital 161 — (161) It is necessary to clarify the responsibilities and competences at Union and national level as regards A…
Recital 162 — (162) To make best use of the centralised Union expertise and synergies at Union level, the powers of supervis…
Recital 163 — (163) With a view to complementing the governance systems for general-purpose AI models, the scientific panel …
Recital 164 — (164) The AI Office should be able to take the necessary actions to monitor the effective implementation of an…
Recital 165 — (165) The development of AI systems other than high-risk AI systems in accordance with the requirements of thi…
Recital 166 — (166) It is important that AI systems related to products that are not high-risk in accordance with this Regul…
Recital 167 — (167) In order to ensure trustful and constructive cooperation of competent authorities on Union and national …
Recital 168 — (168) Compliance with this Regulation should be enforceable by means of the imposition of penalties and other …
Recital 169 — (169) Compliance with the obligations on providers of general-purpose AI models imposed under this Regulation …
Recital 170 — (170) Union and national law already provide effective remedies to natural and legal persons whose rights and …
Recital 171 — (171) Affected persons should have the right to obtain an explanation where a deployer’s decision is based mai…
Recital 172 — (172) Persons acting as whistleblowers on the infringements of this Regulation should be protected under the U…
Recital 173 — (173) In order to ensure that the regulatory framework can be adapted where necessary, the power to adopt acts…
Recital 174 — (174) Given the rapid technological developments and the technical expertise required to effectively apply thi…
Recital 175 — (175) In order to ensure uniform conditions for the implementation of this Regulation, implementing powers sho…
Recital 176 — (176) Since the objective of this Regulation, namely to improve the functioning of the internal market and to …
Recital 177 — (177) In order to ensure legal certainty, ensure an appropriate adaptation period for operators and avoid disr…
Recital 178 — (178) Providers of high-risk AI systems are encouraged to start to comply, on a voluntary basis, with the rele…
Recital 179 — (179) This Regulation should apply from 2 August 2026. However, taking into account the unacceptable risk asso…
Recital 180 — (180) The European Data Protection Supervisor and the European Data Protection Board were consulted in accorda…