Recitals The Artifical Intelligence Act (AI Act) was adopted in consideration of the following recitals. (1)Purpose of the regulation(2)Regulation in balance with the values of the Union(3)Consistent and high level of protection(4)Artificial Intelligence(5)Risks and harms(6)Human-centric technology(7)High-risk AI systems(8)Framework for harmonised rules(9)Application of existing rights and remedies(10)Fundamental right to the protection of personal data(11)Liability of intermediary service providers of Directive 2000/31/EG(12)AI system(13)Deployer(14)Biometric data(15)Biometric identification(16)Biometric categorisation(17)Remote biometric identification system(18)Emotion recognition system(19)Publicly accessible space(20)Equipping with necessary notions(21)Application to providers and deployers(22)Application to third country operators(23)Application to Union institutions, bodies and other agencies(24)No application for military, defence or national security(25)No application to scientific research and development(26)Risk-based approach(27)Principles for trustworthy and ethical AI(28)Prohibition of harmful and misused practices(29)AI-enabled influence on human behaviour(30)AI systems for biometric categorisation(31)AI systems for social scoring(32)AI systems for 'real-time' remote biometric identification(33)Real-time' remote biometric identification systems for law enforcement purposes(34)Fundamental rights impact assessment(35)Authorisation obligation(36)Notification of the authorities(37)Adoption into national law(38)Use and processing of biometric data(39)Applicability of Directive (EU) 2016/680(40)Limited binding effect for Ireland(41)Limited binding effect for Denmark(42)Presumption of innocence in profiling(43)Banning AI facial recognition for mass surveillance(44)Risks and limits of emotion-recognising AI systems(45)Practices not permitted under EU law remain unaffected(46)Mandatory security requirements for high-risk AI systems in the EU(47)Risk minimisation in AI systems for health and safety(48)Fundamental rights and risk assessment of AI systems(49)Adaptation of regulations for high-risk AI systems in safety-critical products and systems(50)High-risk classification for AI security products(51)High-risk AI systems and product risk(52)Classification and regulation of high-risk AI systems(53)Supplementing human activity and pattern recognition(54)Risk assessment and use cases for biometric systems(55)Classification of high-risk AI systems in critical infrastructure(56)High-risk AI systems in education(57)High-risk AI systems in employment and human resources(58)High-risk AI systems deployment in social benefits, creditworthiness, and emergencies(59)High-risk AI systems in law enforcement(60)High-risk AI systems in migration and asylum(61)High-risk AI systems in the administration of justice(62)High-risk AI systems in elections(63)Lawfulness(64)Compliance with other Union law(65)Risk-management system(66)Scope of the requirements(67)Data governance and management practices(68)Access to high-quality data(69)Data privacy(70)Processing of special categories of personal data(71)Documentation of the development(72)Transparency for high-risk AI systems(73)Oversight by natural persons(74)Performance metrics(75)Resilience of high-risk AI systems(76)Cybersecurity appropriate to the risks(77)Cybersecurity requirements as set out in the regulation(78)Conformity assessment(79)Specification of the provider(80)Protection and equality of persons with disabilities(81)Establishing a quality management system(82)Level-playing field for operators(83)Clarification of the roles and obligations of operators along the value chain(84)Extension of obligations to distributors, importers, deployers or other third parties as providers of high-risk AI system(85)Cooperation of high-risk AI system providers along the value chain and with authorities(86)Obligation for former providers to cooperate(87)Obligations in the event of high-risk AI system as a safety component of a product(88)Written agreement on the provision of information(89)Release from value chain responsibilities and promotion of trustworthy AI(90)Development of voluntary model contract terms by the Commission(91)Definition of special responsibilities and obligations for deployers(92)Duty of employers to inform and consult employees or their representatives(93)Notification and information of persons by the deployer(94)Application of Directive (EU) 2016/680 when using AI systems for biometric identification for law enforcement purposes(95)Compliance with safeguards for subsequent remote biometric identification(96)Implementation of fundamental rights impact assessment and further follow-up measures(97)Precise definition of individual AI systems(98)Model determination based on parameters(99)Example of AI model with general-purpose(100)Further cases of AI models with a general-purpose(101)Transparency obligations for providers of general-purpose AI models(102)Transparency obligations for general-purpose AI models under open-source license(103)Defintion of the term AI model with general-purpose under open-source license(104)Exemptions for transparency obligations for general-purpose AI models under open-source license(105)Use of protected content for developing general-purpose AI models(106)Strategy for compliance with Union law for providers of general-purpose AI models(107)Transparency obligations for providers of general-purpose AI models concerning training data(108)Compliance monitoring by the AI Office(109)Proportionality of monitoring measures by the AI Office(110)Systemic risks of general-purpose AI models(111)Methodology for classifying AI models as systemically risky(112)Notification of general-purpose AI models with systemic risk at AI Office(113)Subsequent classificaton of AI models with systemic risk(114)Obligations for providers of general-purpose AI models with systemic risks(115)Mandatory reporting of risk events for providers of general-purpose AI models with systemic risks(116)Codes of practice for providers of general-purpose AI models with systemic risks(117)Implementation of harmonised codes of practice by the EU Commission(118)AI Act and Digital Services Act when embedding AI models in platforms and search engines(119)AI systems as intermediary services under the Digital Services Act(120)Obligation to identify systemic risks for AI models and intermediary services(121)Harmonised standards of the Commission(122)Compliance with cybersecurity requirements(123)Conformity assessment for high-risk AI systems(124)Conformity assessment according to relevant EU harmonisation legislation(125)Limited conformity assessment by third parties(126)Procedure for third-party conformity assessments(127)Recognition of conformity results from third countries(128)Reassessment of conformity in the event of significant system changes(129)CE marking for high-risk AI systems(130)Authorisation for high-risk AI systems in the public interest(131)Registration obligations for providers and deployers of high-risk AI systems in EU database(132)Transparency obligations for AI systems(133)Mandatory labeling of AI results(134)Obligation to disclose artificially generated content(135)Development of codes of conduct at Union level for AI content review(136)Importance of transparency for the implementation of the AI Regulation(137)Transparency is no substitute for lawfulness(138)Establishing of AI regulatory sandboxes to test innovative systems(139)Objectives of AI regulatory sandboxes(140)Legal basis for the use of personal data for the development of AI systems in the public interest(141)Requirements for testing high-risk AI systems under real-world conditions(142)Encourage the funding of research and development on AI solutions(143)Special consideration of the interests of SMEs(144)Promotion and protection of innovation(145)Minimising risk and facilitating the fulfillment of obligations(146)Facilitating the introduction of quality management for microenterprises(147)Facilitating access to testing and experimentation facilites for certain entities(148)Establishment of governance framework(149)Establishment of Regulation Board(150)Establishment of advisory forum(151)Establishment of scientific panel(152)Establishment of Union structures to support the testing of AI(153)Establishment of at least one notifying authority and one market surveillance authority per Member State(154)Independence, impartiality and confidentiality of the relevant national authorities(155)System for monitoring high-risk AI systems and system for reporting serious incidents(156)System of market surveillance and conformity of products(157)Position of national authorities or other bodies in relation to the regulations of the AI Act(158)Coherent application and enforcement of obligations for financial services(159)Necessary powers of market surveillance authorities for high-risk AI systems in the area of biometric data(160)Joint activities of the market surveillance authorities of the Member States or the Commission(161)Clarification of responsibilities and competencies(162)Synergies at Union level(163)Support for the AI Office by scientific panel(164)Competencies of the AI Office(165)Voluntary codes of conduct(166)Safety of AI systems that are not high-risk(167)Trustful cooperation of competent authorities(168)Ensuring the implementation of the provisions(169)Administrative fines(170)Legal remedies for natural and legal persons(171)Explanation on the use of high-risk systems(172)Protection of whistleblowers(173)Delegation of powers to the Commission(174)Regular evaluation and review(175)Implementing powers of the Commission(176)Principle of subsidiarity and principle of proportionality(177)Ensuring continuity(178)Voluntary compliance(179)Date of validity(180)Adoption of the regulation AI Act Table of contents Report error
