Legal text

EU AI Act

Regulation (EU) 2024/1689 — Artificial Intelligence Act

Official source ↗

Article

• Article 1Subject matter
• Article 2Scope
• Article 3Definitions
• Article 4AI literacy
• Article 4aProcessing of special categories of personal data for bias detection and correction
• Article 5Prohibited AI practices
• Article 6Classification rules for high-risk AI systems
• Article 7Amendments to Annex III
• Article 8Compliance with the requirements
• Article 9Risk management system
• Article 10Data and data governance
• Article 11Technical documentation
• Article 12Record-keeping
• Article 13Transparency and provision of information to deployers
• Article 14Human oversight
• Article 15Accuracy, robustness and cybersecurity
• Article 16Obligations of providers of high-risk AI systems
• Article 17Quality management system
• Article 18Documentation keeping
• Article 19Automatically generated logs
• Article 20Corrective actions and duty of information
• Article 21Cooperation with competent authorities
• Article 22Authorised representatives of providers of high-risk AI systems
• Article 23Obligations of importers
• Article 24Obligations of distributors
• Article 25Responsibilities along the AI value chain
• Article 26Obligations of deployers of high-risk AI systems
• Article 27Fundamental rights impact assessment for high-risk AI systems
• Article 28Notifying authorities
• Article 29Application of a conformity assessment body for notification
• Article 30Notification procedure
• Article 31Requirements relating to notified bodies
• Article 32Presumption of conformity with requirements relating to notified bodies
• Article 33Subsidiaries of notified bodies and subcontracting
• Article 34Operational obligations of notified bodies
• Article 35Identification numbers and lists of notified bodies
• Article 36Changes to notifications
• Article 37Challenge to the competence of notified bodies
• Article 38Coordination of notified bodies
• Article 39Conformity assessment bodies of third countries
• Article 40Harmonised standards and standardisation deliverables
• Article 41Common specifications
• Article 42Presumption of conformity with certain requirements
• Article 43Conformity assessment
• Article 44Certificates
• Article 45Information obligations of notified bodies
• Article 46Derogation from conformity assessment procedure
• Article 47EU declaration of conformity
• Article 48CE marking
• Article 49Registration
• Article 50Transparency obligations for providers and deployers of certain AI systems
• Article 51Classification of general-purpose AI models as general-purpose AI models with systemic risk
• Article 52Procedure
• Article 53Obligations for providers of general-purpose AI models
• Article 54Authorised representatives of providers of general-purpose AI models
• Article 55Obligations of providers of general-purpose AI models with systemic risk
• Article 56Codes of practice
• Article 57AI regulatory sandboxes
• Article 58Detailed arrangements for, and functioning of, AI regulatory sandboxes
• Article 59Further processing of personal data for developing certain AI systems in the public interest in the AI regulatory sandbox
• Article 60Testing of high-risk AI systems in real world conditions outside AI regulatory sandboxes
• Article 60aTesting of high-risk AI systems covered by Union harmonisation legislation listed in Section B of Annex I in real-world conditions outside AI regulatory sandboxes
• Article 61Informed consent to participate in testing in real world conditions outside AI regulatory sandboxes
• Article 62Measures for providers and deployers, in particular SMEs, including start-ups
• Article 63Derogations for specific operators
• Article 64AI Office
• Article 65Establishment and structure of the European Artificial Intelligence Board
• Article 66Tasks of the Board
• Article 67Advisory forum
• Article 68Scientific panel of independent experts
• Article 69Access to the pool of experts by the Member States
• Article 70Designation of national competent authorities and single points of contact
• Article 71EU database for high-risk AI systems listed in Annex III
• Article 72Post-market monitoring by providers and post-market monitoring plan for high-risk AI systems
• Article 73Reporting of serious incidents
• Article 74Market surveillance and control of AI systems in the Union market
• Article 75Market surveillance and control of AI systems and mutual assistance
• Article 75aSupervisory and enforcement powers of the AI Office
• Article 75bCommitments
• Article 75cNon-compliance, fines and periodic penalty payments
• Article 75dSafeguards and further specification
• Article 76Supervision of testing in real world conditions by market surveillance authorities
• Article 77Powers of authorities protecting fundamental rights and cooperation with market surveillance authorities
• Article 78Confidentiality
• Article 79Procedure at national level for dealing with AI systems presenting a risk
• Article 80Procedure for dealing with AI systems classified by the provider as non-high-risk in application of Annex III
• Article 81Union safeguard procedure
• Article 82Compliant AI systems which present a risk
• Article 83Formal non-compliance
• Article 84Union AI testing support structures
• Article 85Right to lodge a complaint with a market surveillance authority
• Article 86Right to explanation of individual decision-making
• Article 87Reporting of infringements and protection of reporting persons
• Article 88Enforcement of the obligations of providers of general-purpose AI models
• Article 89Monitoring actions
• Article 90Alerts of systemic risks by the scientific panel
• Article 91Power to request documentation and information
• Article 92Power to conduct evaluations
• Article 93Power to request measures
• Article 94Procedural rights of economic operators of the general-purpose AI model
• Article 95Codes of conduct for voluntary application of specific requirements
• Article 96Guidelines from the Commission on the implementation of this Regulation
• Article 97Exercise of the delegation
• Article 98Committee procedure
• Article 99Penalties
• Article 100Administrative fines on Union institutions, bodies, offices and agencies
• Article 101Fines for providers of general-purpose AI models
• Article 102Amendment to Regulation (EC) No 300/2008
• Article 103Amendment to Regulation (EU) No 167/2013
• Article 104Amendment to Regulation (EU) No 168/2013
• Article 105Amendment to Directive 2014/90/EU
• Article 106Amendment to Directive (EU) 2016/797
• Article 107Amendment to Regulation (EU) 2018/858
• Article 108Amendments to Regulation (EU) 2018/1139
• Article 109Amendment to Regulation (EU) 2019/2144
• Article 110Amendment to Directive (EU) 2020/1828
• Article 111AI systems already placed on the market or put into service and general-purpose AI models already placed on the marked
• Article 112Evaluation and review
• Article 113Entry into force and application

Recital

• Recital 1
• Recital 2
• Recital 3
• Recital 4
• Recital 5
• Recital 6
• Recital 7
• Recital 8
• Recital 9
• Recital 10
• Recital 11
• Recital 12
• Recital 13
• Recital 14
• Recital 15
• Recital 16
• Recital 17
• Recital 18
• Recital 19
• Recital 20
• Recital 21
• Recital 22
• Recital 23
• Recital 24
• Recital 25
• Recital 26
• Recital 27
• Recital 28
• Recital 29
• Recital 30
• Recital 31
• Recital 32
• Recital 33
• Recital 34
• Recital 35
• Recital 36
• Recital 37
• Recital 38
• Recital 39
• Recital 40
• Recital 41
• Recital 42
• Recital 43
• Recital 44
• Recital 45
• Recital 46
• Recital 47
• Recital 48
• Recital 49
• Recital 50
• Recital 51
• Recital 52
• Recital 53
• Recital 54
• Recital 55
• Recital 56
• Recital 57
• Recital 58
• Recital 59
• Recital 60
• Recital 61
• Recital 62
• Recital 63
• Recital 64
• Recital 65
• Recital 66
• Recital 67
• Recital 68
• Recital 69
• Recital 70
• Recital 71
• Recital 72
• Recital 73
• Recital 74
• Recital 75
• Recital 76
• Recital 77
• Recital 78
• Recital 79
• Recital 80
• Recital 81
• Recital 82
• Recital 83
• Recital 84
• Recital 85
• Recital 86
• Recital 87
• Recital 88
• Recital 89
• Recital 90
• Recital 91
• Recital 92
• Recital 93
• Recital 94
• Recital 95
• Recital 96
• Recital 97
• Recital 98
• Recital 99
• Recital 100
• Recital 101
• Recital 102
• Recital 103
• Recital 104
• Recital 105
• Recital 106
• Recital 107
• Recital 108
• Recital 109
• Recital 110
• Recital 111
• Recital 112
• Recital 113
• Recital 114
• Recital 115
• Recital 116
• Recital 117
• Recital 118
• Recital 119
• Recital 120
• Recital 121
• Recital 122
• Recital 123
• Recital 124
• Recital 125
• Recital 126
• Recital 127
• Recital 128
• Recital 129
• Recital 130
• Recital 131
• Recital 132
• Recital 133
• Recital 134
• Recital 135
• Recital 136
• Recital 137
• Recital 138
• Recital 139
• Recital 140
• Recital 141
• Recital 142
• Recital 143
• Recital 144
• Recital 145
• Recital 146
• Recital 147
• Recital 148
• Recital 149
• Recital 150
• Recital 151
• Recital 152
• Recital 153
• Recital 154
• Recital 155
• Recital 156
• Recital 157
• Recital 158
• Recital 159
• Recital 160
• Recital 161
• Recital 162
• Recital 163
• Recital 164
• Recital 165
• Recital 166
• Recital 167
• Recital 168
• Recital 169
• Recital 170
• Recital 171
• Recital 172
• Recital 173
• Recital 174
• Recital 175
• Recital 176
• Recital 177
• Recital 178
• Recital 179
• Recital 180
• Omnibus recital 1
• Omnibus recital 2
• Omnibus recital 3
• Omnibus recital 4
• Omnibus recital 5
• Omnibus recital 6
• Omnibus recital 7
• Omnibus recital 8
• Omnibus recital 9
• Omnibus recital 10
• Omnibus recital 11
• Omnibus recital 12
• Omnibus recital 13
• Omnibus recital 14
• Omnibus recital 15
• Omnibus recital 16
• Omnibus recital 17
• Omnibus recital 18
• Omnibus recital 19
• Omnibus recital 20
• Omnibus recital 21
• Omnibus recital 22
• Omnibus recital 23
• Omnibus recital 24
• Omnibus recital 25
• Omnibus recital 26
• Omnibus recital 27
• Omnibus recital 28
• Omnibus recital 29
• Omnibus recital 30
• Omnibus recital 31
• Omnibus recital 32
• Omnibus recital 33
• Omnibus recital 34
• Omnibus recital 35
• Omnibus recital 36
• Omnibus recital 37
• Omnibus recital 38
• Omnibus recital 39
• Omnibus recital 40
• Omnibus recital 41
• Omnibus recital 42
• Omnibus recital 43
• Omnibus recital 44
• Omnibus recital 45
• Omnibus recital 46
• Omnibus recital 47

Annex

• Annex XIVThe list of codes, categories and corresponding types of AI systems for the purpose of the notification procedure referred to in Article 30 specifying the scope of the designation as notified bodies
• Annex IList of Union harmonisation legislation
• Annex IIList of criminal offences referred to in Article 5(1), first subparagraph, point (h)(iii)
• Annex IIIHigh-risk AI systems referred to in Article 6(2)
• Annex IVTechnical documentation referred to in Article 11(1)
• Annex VEU declaration of conformity
• Annex VIConformity assessment procedure based on internal control
• Annex VIIConformity based on an assessment of the quality management system and an assessment of the technical documentation
• Annex VIIIInformation to be submitted upon the registration of high-risk AI systems in accordance with Article 49
• Annex IXInformation 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
• Annex XUnion legislative acts on large-scale IT systems in the area of Freedom, Security and Justice
• Annex XITechnical documentation referred to in Article 53(1), point (a) — technical documentation for providers of general-purpose AI models
• Annex XIITransparency 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
• Annex XIIICriteria for the designation of general-purpose AI models with systemic risk referred to in Article 51

Definition

• ‘AI system’AI system
• ‘risk’risk
• ‘provider’provider
• ‘deployer’deployer
• ‘authorised representative’authorised representative
• ‘importer’importer
• ‘distributor’distributor
• ‘operator’operator
• ‘placing on the market’placing on the market
• ‘making available on the market’making available on the market
• ‘putting into service’putting into service
• ‘intended purpose’intended purpose
• ‘reasonably foreseeable misuse’reasonably foreseeable misuse
• ‘safety component’safety component
• ‘instructions for use’instructions for use
• ‘recall of an AI system’recall of an AI system
• ‘withdrawal of an AI system’withdrawal of an AI system
• ‘performance of an AI system’performance of an AI system
• ‘notifying authority’notifying authority
• ‘conformity assessment’conformity assessment
• ‘conformity assessment body’conformity assessment body
• ‘notified body’notified body
• ‘substantial modification’substantial modification
• ‘CE marking’CE marking
• ‘post-market monitoring system’post-market monitoring system
• ‘market surveillance authority’market surveillance authority
• ‘harmonised standard’harmonised standard
• ‘common specification’common specification
• ‘training data’training data
• ‘validation data’validation data
• ‘validation data set’validation data set
• ‘testing data’testing data
• ‘input data’input data
• ‘biometric data’biometric data
• ‘biometric identification’biometric identification
• ‘biometric verification’biometric verification
• ‘special categories of personal data’special categories of personal data
• ‘sensitive operational data’sensitive operational data
• ‘emotion recognition system’emotion recognition system
• ‘biometric categorisation system’biometric categorisation system
• ‘remote biometric identification system’remote biometric identification system
• ‘real-time remote biometric identification system’real-time remote biometric identification system
• ‘post-remote biometric identification system’post-remote biometric identification system
• ‘publicly accessible space’publicly accessible space
• ‘law enforcement authority’law enforcement authority
• ‘law enforcement’law enforcement
• ‘AI Office’AI Office
• ‘national competent authority’national competent authority
• ‘serious incident’serious incident
• ‘personal data’personal data
• ‘non-personal data’non-personal data
• ‘profiling’profiling
• ‘real-world testing plan’real-world testing plan
• ‘sandbox plan’sandbox plan
• ‘AI regulatory sandbox’AI regulatory sandbox
• ‘AI literacy’AI literacy
• ‘testing in real-world conditions’testing in real-world conditions
• ‘subject’subject
• ‘informed consent’informed consent
• ‘deep fake’deep fake
• ‘widespread infringement’widespread infringement
• ‘critical infrastructure’critical infrastructure
• ‘general-purpose AI model’general-purpose AI model
• ‘high-impact capabilities’high-impact capabilities
• ‘systemic risk’systemic risk
• ‘general-purpose AI system’general-purpose AI system
• ‘floating-point operation’floating-point operation
• ‘downstream provider’downstream provider

Guidance

• Prohibitions § 1Background and objectives
• Prohibitions § 2Overview of prohibited AI practices
• Prohibitions § 2.1Prohibitions listed in Article 5 AI Act
• Prohibitions § 2.2Legal basis of the prohibitions
• Prohibitions § 2.3Material scope: practices related to the ‘placing on the market’, ‘putting into service’ or ‘use’ of an AI system
• Prohibitions § 2.4Personal scope: responsible actors
• Prohibitions § 2.5Exclusion from the scope of the AI Act
• Prohibitions § 2.5.1National security, defence and military purposes
• Prohibitions § 2.5.2Judicial and law enforcement cooperation with third countries
• Prohibitions § 2.5.3Research & Development
• Prohibitions § 2.5.4Personal non-professional activity
• Prohibitions § 2.5.5AI systems released under free and open source licences
• Prohibitions § 2.6Interplay of the prohibitions with the requirements for high-risk AI systems
• Prohibitions § 2.7Application of the prohibitions to general-purpose AI systems and systems with intended purposes
• Prohibitions § 2.8Interplay between the prohibitions and other Union law
• Prohibitions § 2.9.1Market Surveillance Authorities
• Prohibitions § 2.9.2Penalties
• Prohibitions § 3Article 5(1)(a) and (b) AI Act – harmful manipulation, deception and exploitation
• Prohibitions § 3.1Rationale and objectives
• Prohibitions § 3.2Main components of the prohibition in Article 5(1)(a) AI Act – harmful manipulation
• Prohibitions § 3.2.1Subliminal, purposefully manipulative or deceptive techniques
• Prohibitions § 3.2.2With the objective or the effect of materially distorting the behaviour of a person or a group of persons
• Prohibitions § 3.2.3(Reasonably likely to) cause significant harm
• Prohibitions § 3.3Main components of the prohibition in Article 5(1)(b) AI Act – harmful exploitation of vulnerabilities
• Prohibitions § 3.3.1Exploitation of vulnerabilities due to age, disability , or a specific social or-economic situation
• Prohibitions § 3.3.2With the objective or the effect of materially distorting behaviour
• Prohibitions § 3.3.3(Reasonably likely to) cause significant harm
• Prohibitions § 3.4Interplay between the prohibitions in Article 5(1)(a) and (b) AI Act
• Prohibitions § 3.5Out of scope
• Prohibitions § 3.5.1Lawful persuasion
• Prohibitions § 3.5.2Manipulative, deceptive and exploitative AI systems that are not likely to cause significant harm
• Prohibitions § 3.6Interplay with other Union law
• Prohibitions § 4Article 5(1)(c) AI Act-social scoring
• Prohibitions § 4.1Rationale and objectives
• Prohibitions § 4.2Main concepts and components of the ‘social scoring’ prohibition
• Prohibitions § 4.2.1‘Social scoring’: evaluation or classification based on social behaviour or personal or personality characteristics over a certain period of time
• Prohibitions § 4.2.2The social score must lead to detrimental or unfavourable treatment in unrelated social context s and/or unjustified or disproportionate treatment to the gravity of the social behaviour
• Prohibitions § 4.2.3Regardless of whether provided or used by public or private persons
• Prohibitions § 4.3Out of scope
• Prohibitions § 4.4Interplay with other Union legal acts
• Prohibitions § 5Article 5 ( 1)(d) AI Act – individual risk assessment and prediction of criminal offences
• Prohibitions § 5.1Rationale and objectives
• Prohibitions § 5.2Main concepts and components of the prohibition
• Prohibitions § 5.2.1Assessing the risk or predicting the likelihood of a person committing a crime
• Prohibitions § 5.2.2Solely based on profiling of a natural person or on assessing their personality traits and characteristics
• Prohibitions § 5.2.3Exclusion of AI systems to support the human assessment based on objective and verifiable facts directly linked to a criminal activity
• Prohibitions § 5.2.4Extent to which private actors’ activities may fall within scope
• Prohibitions § 5.3.1Location-based or geospatial predictive or place-based crime predictions
• Prohibitions § 5.3.2AI systems that support human assessments based on objective and verifiable facts linked to a criminal activity
• Prohibitions § 5.3.3AI systems used for crime predictions and assessments in relation to legal entities
• Prohibitions § 5.3.4AI systems used for individual predictions of administrative offences
• Prohibitions § 5.4Interplay with other Union legal acts
• Prohibitions § 6Article 5(1)(e) AI Act-untargeted scraping of facial images
• Prohibitions § 6.1Rationale and objectives
• Prohibitions § 6.2Main concepts and components of the prohibition
• Prohibitions § 6.2.1Facial recognition databases
• Prohibitions § 6.2.2Through untargeted scraping of facial images
• Prohibitions § 6.2.3From the Internet and CCTV footage
• Prohibitions § 6.3Out of scope
• Prohibitions § 6.4Interplay with other Union legal acts
• Prohibitions § 7Article 5(1)(f) AI Act emotion recognition
• Prohibitions § 7.1Rationale and objectives
• Prohibitions § 7.2Main concepts and components of the prohibition
• Prohibitions § 7.2.1AI systems to infer emotions
• Prohibitions § 7.2.2Limitation of the prohibition to workplace and educational institutions
• Prohibitions § 7.2.3Exceptions for medical and safety reasons
• Prohibitions § 7.3More favourable Member State law
• Prohibitions § 7.4Out of scope
• Prohibitions § 8Article 5(1)(g) AI Act: Biometric categorisation for certain ‘sensitive’ characteristics
• Prohibitions § 8.1Rationale and objectives
• Prohibitions § 8.2Main concepts and components of the prohibition
• Prohibitions § 8.2.1Biometric categorisation system
• Prohibitions § 8.2.2Persons are individually categorised based on their biometric data
• Prohibitions § 8.2.3To deduce or infer their race, political opinions, trade union membership, religious or philosophical beliefs, sex-life or sexual orientation
• Prohibitions § 8.3Out of scope
• Prohibitions § 8.4Interplay with other Union law
• Prohibitions § 9Article 5(1)(h) AI Act-Real-time Remote Biometric Identification (rbi) Systems for Law Enforcement Purposes
• Prohibitions § 9.1Rationale and objectives
• Prohibitions § 9.2Main concepts and components of the prohibition
• Prohibitions § 9.2.1The Notion of Remote Biometric Identification
• Prohibitions § 9.2.2Real-time
• Prohibitions § 9.2.3In publicly accessible spaces
• Prohibitions § 9.2.4For law enforcement purposes
• Prohibitions § 9.3Exceptions to the prohibition
• Prohibitions § 9.3.1Rationale and objectives
• Prohibitions § 9.3.2Targeted search for the victims of three serious crimes and missing persons
• Prohibitions § 9.3.3Prevention of imminent threats to life or terrorist attacks
• Prohibitions § 9.3.4Localisation and identification of suspects of certain crimes
• Prohibitions § 10.1Targeted individual and safeguards (Article 5(2) AI Act)
• Prohibitions § 10.1.1Fundamental Rights Impact Assessment
• Prohibitions § 10.1.2Registration of the authorized RBI systems
• Prohibitions § 10.2Need for prior authorisation
• Prohibitions § 10.2.1Objective
• Prohibitions § 10.2.2The main principle: Prior authorisation by a judicial authority or an independent administrative authority
• Prohibitions § 10.2.2.1Prior and reasoned request in accordance with national procedural rules
• Prohibitions § 10.2.2.2Authorisation by a judicial authority or an independent administrative authority
• Prohibitions § 10.2.2.3The exception to the requirement of prior authorisation: request within 24 hours and consequences if rejected
• Prohibitions § 10.2.2.4Immediate cessation in case the request for authorisation is rejected and deletion of the data
• Prohibitions § 10.2.2.5No decision-making solely on the output of the real-time RBI system
• Prohibitions § 10.3Notification to the authorities of each use of ‘real-time’ remote biometric identification systems in publicly accessible spaces for law enforcement
• Prohibitions § 10.4.1Principle: national law required to provide the legal basis for the authorisation for all or some of the exceptions
• Prohibitions § 10.4.2National law shall respect the limits and conditions of Article 5(1)(h) AI Act
• Prohibitions § 10.4.3Detailed national law on the authorisation request, the issuance and the exercise
• Prohibitions § 10.4.4Detailed national law on the supervision and the reporting relating to the authorisation
• Prohibitions § 10.5Annual reports by the national market surveillance authorities and the national data protection authorities of Member States
• Prohibitions § 10.6Annual reports by the Commission
• Prohibitions § 10.7Out-of-Scope
• Prohibitions § 10.8Examples of uses
• Prohibitions § 11Entry into application
• Prohibitions § 12Review and update of the Commission guidelines
• High-risk § IIntroduction
• High-risk § IIGeneral principles for classification of high-risk AI systems
• High-risk § II.1The system must be an AI system
• High-risk § II.2Intended purpose(s) of the AI system
• High-risk § 1Rationale, approach
• High-risk § 2Horizontal issues for Annex III use cases
• High-risk § 2.1The role of human involvement and high-risk classification
• High-risk § 2.2Limitations in some use cases to natural persons
• High-risk § 2.3AI systems as part of complex systems/products and services
• High-risk § 2.4‘Intended to be used’
• High-risk § 2.5‘On behalf of’
• High-risk § 2.6‘In so far as their use is permitted under relevant Union or national law’
• High-risk § 2.7‘Filter’ to exempt AI systems from being high-risk according to Article 6(3)
• High-risk § 2.7.1Conditions for the ‘filter’ to apply
• High-risk § 2.7.2Exception for profiling
• High-risk § 2.7.3Self-assessment by the provider, registration and monitoring of systems benefitting from the filter mechanism
• High-risk § 2.7.4Protection against circumvention
• High-risk § 3.1Biometrics
• High-risk § 3.1.1Overview of use cases and horizontal issues
• High-risk § 3.1.2Point 1(a): Remote biometric identification systems
• High-risk § 3.1.2.1Remote biometric identification system
• High-risk § 3.1.2.2Authorisation requirement for certain law enforcement uses of post RBI under Article 26(10) AI Act
• High-risk § 3.1.3Point 1(b): Biometric categorisation
• High-risk § 3.1.4Point 1(c): Emotion recognition
• High-risk § 3.2Critical infrastructure
• High-risk § 3.2.1Overview of use cases and horizontal issues
• High-risk § 3.2.2The critical digital infrastructure use case
• High-risk § 3.2.3The road traffic use case in critical infrastructure
• High-risk § 3.2.4The supply of water use case in critical infrastructure
• High-risk § 3.2.5The supply of gas use case in critical infrastructure
• High-risk § 3.2.6The supply of heating use case in critical infrastructure
• High-risk § 3.2.7The supply of electricity use case in critical infrastructure
• High-risk § 3.3Education and vocational training
• High-risk § 3.3.1Overview of use cases and horizontal issues
• High-risk § 3.3.2Point 3(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
• High-risk § 3.3.3Point 3(b): AI systems intended to be used to evaluate learning outcomes
• High-risk § 3.3.4Point 3(c): AI systems intended to be used for the purpose of assessing the appropriate level of education
• High-risk § 3.3.5Point 3(d): AI systems intended to be used for monitoring and detecting prohibited behaviour of students
• High-risk § 3.4Employment
• High-risk § 3.4.1Horizontal issues
• High-risk § 3.4.2Point 4(a): AI systems intended to be used for the recruitment or selection of natural persons
• High-risk § 3.4.3Point 4(b): AI systems intended to be used to manage work-related relationships
• High-risk § 3.5Access to and enjoyment of essential private services and essential public services and benefits
• High-risk § 3.5.1Horizontal issues and overview of use cases
• High-risk § 3.5.2Point 5(a): Evaluation of the eligibility of a natural person for essential public assistance benefits and services, and the granting or denying such benefits and services
• High-risk § 3.5.3Point 5(b): AI systems intended to be used to evaluate the creditworthiness of natural persons or establish their credit score
• High-risk § 3.5.4Point 5(c): AI systems intended to be used for risk assessment and pricing in the case of life and health insurance
• High-risk § 3.5.5Interplay with other Union legislation: Article 144 of Regulation (eu) No 575/2013 (Capital Requirements Regulation) and Article 120 of Directive /2009/13/8 ec (Solvency II Directive)
• High-risk § 3.5.6Point 5(d): Evaluation and classification of emergency calls or the dispatching or establishing prioritisation of emergency first response services
• High-risk § 3.6Law enforcement
• High-risk § 3.6.1Overview of use cases and horizontal issues
• High-risk § 3.6.2Point 6(a): Assessing the risk of a natural person becoming the victim of a criminal offence
• High-risk § 3.6.3Point 6(b): Polygraphs and or similar tools
• High-risk § 3.6.4Point 6(c): AI systems intended to be used to evaluate the reliability of evidence
• High-risk § 3.6.5Point 6(d): AI systems assessing offending or reoffending of concrete person(s)
• High-risk § 3.6.6Point 6(e): AI systems intended to be used for the profiling of natural persons in the course of the detection, investigation or prosecution of criminal offences
• High-risk § 3.6.7Other AI systems falling outside the use cases of point 6 of Annex III
• High-risk § 3.7Migration, asylum and border control management
• High-risk § 3.7.1Overview of use cases and horizontal issues
• High-risk § 3.7.2Point 7(a): AI systems intended to be used as polygraphs or similar tools
• High-risk § 3.7.3Point 7(b): AI systems intended to be used for risk assessment of persons seeking to enter or stay
• High-risk § 3.7.4Point 7(c): AI systems intended to be used for assistance in examining asylum/visa/residence applications and associated complaints
• High-risk § 3.7.5Point 7(d): AI systems intended to be used for detecting, recognising or identifying natural persons in migration, asylum or border control management contexts (excluding travel-document verification)
• High-risk § 3.8Administration of justice and democratic processes
• High-risk § 3.8.1Point 8(a): AI systems intended to be used to assist judicial authorities or in alternative dispute resolution
• High-risk § 3.8.2Point 8 (b): AI systems intended to be used for influencing the outcome of elections or referendum
• High-risk § VEntry into application of the rules for high-risk AI systems
• High-risk § VIReview and update of the high-risk use cases and the Commission guidelines
• Transparency § 1Background and objectives
• Transparency § 2.1Overview of the transparency obligations in Article 50 AI Act
• Transparency § 2.2Rationale and objectives
• Transparency § 2.3Responsible actors under Article 50 AI Act
• Transparency § 2.4Exclusions from the scope of the AI Act
• Transparency § 2.4.1Deployers acting in the context of purely personal non-professional activity
• Transparency § 2.4.2Research & Development
• Transparency § 2.4.3AI released under free and open-source licences
• Transparency § 2.5Interplay with the prohibited practices, the requirements for high-risk AI systems and AI literacy
• Transparency § 2.6Interplay with rules applicable to general-purpose AI models and/or systems
• Transparency § 3Article 50(1) AI Act: transparency for interactive AI systems
• Transparency § 3.1Main components, concepts and related transparency obligation(s)
• Transparency § 3.1.1AI systems intended to interact directly with natural persons
• Transparency § 3.1.2Information obligation under Article 50(1) AI Act
• Transparency § 3.2Exceptions to the information obligation under Article 50(1) AI Act
• Transparency § 3.2.1Exception for obvious interaction with an AI system
• Transparency § 3.2.2Exception for AI systems authorised by law for law enforcement purposes
• Transparency § 3.3Interplay with other Union legal acts
• Transparency § 4Article 50(2) AI Act: marking and detection of ai-generated or manipulated content
• Transparency § 4.1Main components and concepts of Article 50(2) AI Act
• Transparency § 4.1.1AI systems generating or manipulating synthetic content
• Transparency § 4.1.2Modalities of synthetic content in scope
• Transparency § 4.1.3Content falling outside the scope of Article 50(2) AI Act
• Transparency § 4.2The marking and detection obligation of Article 50(2) AI Act
• Transparency § 4.2.1The marking obligation
• Transparency § 4.2.2The detection obligation
• Transparency § 4.2.3Compliance with the requirements for technical solutions: effective, interoperable, robust and reliable
• Transparency § 4.3Exceptions to the obligations under Article 50(2) AI Act
• Transparency § 4.4Interplay with other Union legal acts
• Transparency § 5Article 50(3) AI Act: emotion recognition systems and biometric categorisation systems
• Transparency § 5.1.1The notion of an emotion recognition system
• Transparency § 5.1.2The notion of a biometric categorisation system
• Transparency § 5.1.3The information obligation under Article 50(3) AI Act
• Transparency § 5.2Exception for law enforcement purposes
• Transparency § 5.3Interplay with other Union legal acts
• Transparency § 6Article 50(4): labelling of deep fakes and certain text publications
• Transparency § 6.1Main components, concepts and related transparency obligation(s) for deep fakes under Article 50(4) AI Act
• Transparency § 6.1.1The notion of ‘deep fake’
• Transparency § 6.1.2The disclosure obligation under Article 50(4), first subparagraph, AI Act
• Transparency § 6.1.3Transparency of artistic, creative, satirical, fictional or analogous deep fake content
• Transparency § 6.1.4Exception for law enforcement
• Transparency § 6.1.5Interplay with other Union legal acts
• Transparency § 6.2Main components, concepts and related transparency obligation(s) for AI-generated or manipulated text under Article 50(4) AI Act
• Transparency § 6.2.1Text published with the purpose of informing the public on matters of public interest
• Transparency § 6.2.2The disclosure obligation under Article 50(4), second subparagraph, AI Act
• Transparency § 6.2.3Exception from the transparency obligation for text under human review or editorial control and editorial responsibility
• Transparency § 6.2.4Exception for law enforcement
• Transparency § 6.2.5Interplay with other Union legal acts
• Transparency § 7Horizontal requirements applicable to the information provided under article 50(5) AI Act
• Transparency § 7.1Information provision in a clear and distinguishable manner
• Transparency § 7.2First interaction or exposure
• Transparency § 7.3Compliance with applicable accessibility requirements
• Transparency § 8Enforcement of article 50 AI Act
• Transparency § 8.1Effects of adhering to a code of practice assessed as adequate
• Transparency § 8.2Market Surveillance Authorities
• Transparency § 8.3Penalties
• Transparency § 8.4Entry into application
• Transparency § 9Review and update of the guidelines