- 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 4a
Processing of special categories of personal data for bias detection and correction
To the extent strictly necessary to ensure bias detection and correction in relation to high-risk AI systems in accordance with Article 10(2), points (f) and (g), of this Regulation, 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, as applicable, all the following conditions shall be met in order for such processing to occur:
the bias detection and correction cannot be effectively fulfilled by processing other data, including synthetic or anonymised data;
the special categories of personal data are subject to technical limitations on the re-use of personal data, and state-of-the-art security and privacy-preserving measures, including pseudonymisation;
the special categories of personal data are subject to measures to ensure that the personal data processed are secured and protected, subject to suitable safeguards, including strict controls and documentation of the access, to avoid misuse and to ensure that only authorised persons have access to those personal data with appropriate confidentiality obligations;
the special categories of personal data are not transmitted, transferred or otherwise accessed by other parties;
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; and
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.
Providers and deployers of other AI systems and models and deployers of high-risk AI systems may exceptionally process special categories of personal data to the extent that:
such processing is strictly necessary to ensure bias detection and correction 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 pursuant to Union law, especially where data outputs influence inputs for future operations; and
all of the conditions and safeguards set out in paragraph 1 are applied.
This paragraph does not create any obligation to conduct such bias detection and correction.
