- Art. 6Classification rules for high-risk AI systems
- Art. 7Amendments to Annex III
- Art. 8Compliance with the requirements
- Art. 9Risk management system
- Art. 10Data and data governance
- Art. 11Technical documentation
- Art. 12Record-keeping
- Art. 13Transparency and provision of information to deployers
- Art. 14Human oversight
- Art. 15Accuracy, robustness and cybersecurity
- Art. 16Obligations of providers of high-risk AI systems
- Art. 17Quality management system
- Art. 18Documentation keeping
- Art. 19Automatically generated logs
- Art. 20Corrective actions and duty of information
- Art. 21Cooperation with competent authorities
- Art. 22Authorised representatives of providers of high-risk AI systems
- Art. 23Obligations of importers
- Art. 24Obligations of distributors
- Art. 25Responsibilities along the AI value chain
- Art. 26Obligations of deployers of high-risk AI systems
- Art. 27Fundamental rights impact assessment for high-risk AI systems
- Art. 28Notifying authorities
- Art. 29Application of a conformity assessment body for notification
- Art. 30Notification procedure
- Art. 31Requirements relating to notified bodies
- Art. 32Presumption of conformity with requirements relating to notified bodies
- Art. 33Subsidiaries of notified bodies and subcontracting
- Art. 34Operational obligations of notified bodies
- Art. 35Identification numbers and lists of notified bodies
- Art. 36Changes to notifications
- Art. 37Challenge to the competence of notified bodies
- Art. 38Coordination of notified bodies
- Art. 39Conformity assessment bodies of third countries
- Art. 40Harmonised standards and standardisation deliverables
- Art. 41Common specifications
- Art. 42Presumption of conformity with certain requirements
- Art. 43Conformity assessment
- Art. 44Certificates
- Art. 45Information obligations of notified bodies
- Art. 46Derogation from conformity assessment procedure
- Art. 47EU declaration of conformity
- Art. 48CE marking
- Art. 49Registration
- Art. 51Classification of general-purpose AI models as general-purpose AI models with systemic risk
- Art. 52Procedure
- Art. 53Obligations for providers of general-purpose AI models
- Art. 54Authorised representatives of providers of general-purpose AI models
- Art. 55Obligations of providers of general-purpose AI models with systemic risk
- Art. 56Codes of practice
- Art. 57AI regulatory sandboxes
- Art. 58Detailed arrangements for, and functioning of, AI regulatory sandboxes
- Art. 59Further processing of personal data for developing certain AI systems in the public interest in the AI regulatory sandbox
- Art. 60Testing of high-risk AI systems in real world conditions outside AI regulatory sandboxes
- Art. 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
- Art. 61Informed consent to participate in testing in real world conditions outside AI regulatory sandboxes
- Art. 62Measures for providers and deployers, in particular SMEs, including start-ups
- Art. 63Derogations for specific operators
- Art. 64AI Office
- Art. 65Establishment and structure of the European Artificial Intelligence Board
- Art. 66Tasks of the Board
- Art. 67Advisory forum
- Art. 68Scientific panel of independent experts
- Art. 69Access to the pool of experts by the Member States
- Art. 70Designation of national competent authorities and single points of contact
- Art. 72Post-market monitoring by providers and post-market monitoring plan for high-risk AI systems
- Art. 73Reporting of serious incidents
- Art. 74Market surveillance and control of AI systems in the Union market
- Art. 75Market surveillance and control of AI systems and mutual assistance
- Art. 75aSupervisory and enforcement powers of the AI Office
- Art. 75bCommitments
- Art. 75cNon-compliance, fines and periodic penalty payments
- Art. 75dSafeguards and further specification
- Art. 76Supervision of testing in real world conditions by market surveillance authorities
- Art. 77Powers of authorities protecting fundamental rights and cooperation with market surveillance authorities
- Art. 78Confidentiality
- Art. 79Procedure at national level for dealing with AI systems presenting a risk
- Art. 80Procedure for dealing with AI systems classified by the provider as non-high-risk in application of Annex III
- Art. 81Union safeguard procedure
- Art. 82Compliant AI systems which present a risk
- Art. 83Formal non-compliance
- Art. 84Union AI testing support structures
- Art. 85Right to lodge a complaint with a market surveillance authority
- Art. 86Right to explanation of individual decision-making
- Art. 87Reporting of infringements and protection of reporting persons
- Art. 88Enforcement of the obligations of providers of general-purpose AI models
- Art. 89Monitoring actions
- Art. 90Alerts of systemic risks by the scientific panel
- Art. 91Power to request documentation and information
- Art. 92Power to conduct evaluations
- Art. 93Power to request measures
- Art. 94Procedural rights of economic operators of the general-purpose AI model
- Art. 102Amendment to Regulation (EC) No 300/2008
- Art. 103Amendment to Regulation (EU) No 167/2013
- Art. 104Amendment to Regulation (EU) No 168/2013
- Art. 105Amendment to Directive 2014/90/EU
- Art. 106Amendment to Directive (EU) 2016/797
- Art. 107Amendment to Regulation (EU) 2018/858
- Art. 108Amendments to Regulation (EU) 2018/1139
- Art. 109Amendment to Regulation (EU) 2019/2144
- Art. 110Amendment to Directive (EU) 2020/1828
- Art. 111AI systems already placed on the market or put into service and general-purpose AI models already placed on the marked
- Art. 112Evaluation and review
- Art. 113Entry into force and application
AI Act Article 75a
Supervisory and enforcement powers of the AI Office
Where a market surveillance authority has reason to suspect non-compliance with this Regulation by a provider or a deployer of an AI system referred to in Article 75(1), it may send a request to the AI Office to assess the matter.
When requesting information, the AI Office shall state the legal basis and the purpose of the request, specify what information is required, and set the period within which the information is to be provided. Where the request is a simple request, the AI Office shall additionally indicate that although there is no obligation to provide the information requested, in the case of a voluntary reply, the information must be correct and not misleading, and indicate the potential fines provided for in Article 99(5) for supplying incorrect or misleading information. Where the request is made by decision, the AI Office shall additionally indicate the fines provided for in Article 99(5) for supplying incorrect, incomplete or misleading information and indicate the right to have the decision reviewed by the Court of Justice of the European Union. The AI Office shall send a copy of the request to the market surveillance authority of the Member State in the territory of which the operator or its legal representative is situated.
During such an inspection, the officials of the AI Office shall be empowered to:
enter any of the business premises, land or property located in the Union of the operator concerned;
examine the books, data and other material relevant to the execution of their tasks, irrespective of the medium on which they are stored;
take or obtain in any form copies of or extracts from books, data and other records;
ask any of the persons subject to the inspection, or their representatives, or staff, for oral or written explanations on factors or documents relating to the subject matter and purpose of the inspection, and to record the answers;
seal any business premises and books or records for the duration of, and to the extent necessary for, the inspection.
Where the AI Office finds that a natural or legal person opposes or obstructs an inspection, the national competent authority of the Member State concerned shall afford it the necessary assistance, requesting, where appropriate, the assistance of the police or an equivalent enforcement authority, to enable it to conduct its on-site inspection.
Where an on-site inspection of business premises, land or property requires authorisation by a judicial authority in accordance with national law, the AI Office shall apply for such an authorisation. The AI Office may also apply for such authorisation as a precautionary measure. Where such an authorisation is applied for, the national judicial authority shall promptly verify that the coercive measures envisaged are neither arbitrary nor excessive having regard to the subject matter of the investigation or inspection and the documents provided by the AI Office with the decision. In its verification of the proportionality of coercive measures, the national judicial authority may ask the AI Office for detailed explanations, in particular relating to the grounds the AI Office has for suspecting that an infringement of this Regulation has taken place and the seriousness of the suspected infringement and, where relevant, the nature of the involvement of the person subject to the coercive measures. The national judicial authority shall not review the necessity of the investigation or inspection nor demand information from the case file of the AI Office. In accordance with the Treaties, the legality of the decision of the AI Office is subject to review only by the Court of Justice of the European Union.
In addition to the powers set out in paragraph 1 of this Article, the AI Office, in the exercise of its competences referred to in Article 75(1), may:
order operators to provide access to, and explanations relating to, their AI systems;
impose an obligation on an operator to retain all data and documents deemed to be necessary to assess the implementation of and compliance with the obligations under this Regulation.
