EU AI Act — Official Text (Regulation (EU) 2024/1689)

Regulation (EU) 2024/1689 as 186 rules, each citing the Article or Annex it comes from. The law decides who is high-risk (Article 6) and which models have systemic risk (Articles 51-52) from the facts it reads, never from a self-description. Every rule evaluates in the open verifier, and twelve plain-language intake questions reach a firm answer for a routine case. Read from the official Journal text.

Answers: Classify AI system risk tier, apply prohibited-practice and high-risk obligations, and route governance decisions under the EU AI Act.

Publisher
The Nomos Exchange
Version
1.4.0
Jurisdiction
EU
Rules
186
Verification tier
Compiled
Published
2026-11-14
Price
Free to query

Source: REGULATION (EU) 2024:1689 OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL of 13 June 2024.pdf

Rules

  1. Article 2. The Regulation does not apply to AI systems placed on the market, put into service or used exclusively for military, defence or national security purposes.
  2. Article 2. The Regulation does not apply to AI systems or models developed and put into service for the sole purpose of scientific research and development.
  3. Article 2. The Regulation does not apply to research, testing or development of AI systems or models before they are placed on the market or put into service.
  4. Article 2. The obligations of deployers do not apply to natural persons using AI systems in a purely personal non-professional activity.
  5. Article 2. The Regulation does not apply to AI systems released under free and open-source licences, unless placed on the market or put into service as high-risk, under Article 5 or under Article 50.
  6. Article 4. Providers and deployers must take measures to ensure, to their best extent, a sufficient level of AI literacy of their staff and others operating or using AI systems on their behalf.
  7. Article 5. Placing on the market, putting into service or using an AI system that deploys subliminal, manipulative or deceptive techniques materially distorting behaviour and causing significant harm is prohibited.
  8. Article 5. An AI system that exploits vulnerabilities due to age, disability or social or economic situation, materially distorting behaviour and causing significant harm, is prohibited.
  9. Article 5. Social scoring of persons or groups is prohibited where the score leads to detrimental treatment in unrelated contexts or treatment unjustified or disproportionate to their behaviour.
  10. Article 5. Assessing or predicting the risk of a person committing a crime based solely on profiling or personality traits is prohibited, except to support a human assessment already based on objective, verifiable facts.
  11. Article 5. Creating or expanding facial recognition databases through untargeted scraping of facial images from the internet or CCTV footage is prohibited.
  12. Article 5. Inferring emotions of a person in the workplace or education institutions is prohibited, except for medical or safety reasons.
  13. Article 5. Biometric categorisation inferring race, political opinions, union membership, religious or philosophical beliefs, sex life or sexual orientation is prohibited, except lawful dataset labelling or law-enforcement categorisation.
  14. Article 5. Using 'real-time' remote biometric identification in publicly accessible spaces for law enforcement is prohibited unless strictly necessary for searching victims or missing persons, preventing an imminent threat or terrorist attack, or locating a suspect of a serious offence.
  15. Article 5. Locating or identifying a suspect is a permitted objective only for an offence listed in Annex II and punishable in the Member State concerned by a custodial sentence or detention order of a maximum of at least four years.
  16. Article 5. A permitted use of real-time remote biometric identification must take into account the seriousness, probability and scale of the harm without its use, and its consequences for the rights and freedoms of all persons concerned.
  17. Article 5. A permitted use of real-time remote biometric identification requires a completed fundamental rights impact assessment under Article 27.
  18. Article 5. The system must be registered in the EU database before use, except in duly justified urgency where registration is completed without undue delay.
  19. Article 5. Each use requires prior authorisation by a judicial authority or binding independent administrative authority, except in duly justified urgency.
  20. Article 5. In urgency, use may start without authorisation only if authorisation is requested without undue delay and at the latest within 24 hours.
  21. Article 5. If the authorisation is rejected, the use must stop with immediate effect and all data, results and outputs of that use must be immediately discarded and deleted.
  22. Article 5. No decision that produces an adverse legal effect on a person may be taken based solely on the output of the real-time remote biometric identification system.
  23. Article 5. Each use must be notified to the relevant market surveillance authority and the national data protection authority.
  24. Article 6. An AI system that is a safety component of, or is itself, an Annex I product required to undergo third-party conformity assessment is high-risk.
  25. Article 6. AI systems referred to in Annex III are high-risk.
  26. Article 6. An Annex III system is not high-risk where it performs a narrow procedural task, improves a completed human activity, detects decision patterns without replacing human assessment, or performs a preparatory task.
  27. Article 6. An Annex III system that performs profiling of natural persons is always high-risk, notwithstanding the derogation.
  28. Article 6. A provider that considers an Annex III system not high-risk must document its assessment before placing it on the market or putting it into service, and is subject to registration under Article 49(2).
  29. Article 8. High-risk AI systems must comply with the requirements of Chapter III, Section 2, taking into account their intended purpose and the state of the art.
  30. Article 8. Where a product contains an AI system subject to this Regulation and Annex I Section A legislation, the provider must ensure the product fully complies with all applicable requirements.
  31. Article 9. A risk management system must be established, implemented, documented and maintained as a continuous iterative process identifying, estimating and evaluating risks and adopting targeted measures.
  32. Article 9. Risk management measures must leave residual risks judged acceptable, by eliminating or reducing risks through design, mitigating what remains and informing and training deployers.
  33. Article 9. High-risk AI systems must be tested against prior defined metrics and probabilistic thresholds, at the latest before being placed on the market or put into service.
  34. Article 9. Providers must consider whether the system is likely to adversely impact persons under 18 and other vulnerable groups.
  35. Article 10. Systems trained with data must be developed on data sets that are relevant, sufficiently representative, as error-free and complete as possible, and suited to the setting of use.
  36. Article 10. Data sets must be subject to governance practices covering design choices, data origin, preparation, assumptions, availability, bias examination and mitigation, and data gaps.
  37. Article 10. Special categories of personal data may be processed for bias detection and correction only where strictly necessary and all the Article 10(5) conditions are met.
  38. Article 11. Technical documentation demonstrating compliance must be drawn up before placing on the market or putting into service and kept up to date; one set for Annex I Section A products.
  39. Article 12. High-risk AI systems must technically allow automatic logging over their lifetime, enabling identification of risks, post-market monitoring and monitoring of operation.
  40. Article 12. Annex III point 1(a) systems must log each period of use, the reference database, matched input data and the persons verifying results.
  41. Article 13. High-risk AI systems must be designed so that their operation is sufficiently transparent for deployers to interpret the output and use it appropriately.
  42. Article 13. High-risk AI systems must be accompanied by concise, complete, correct and clear instructions for use containing at least the information listed in Article 13(3).
  43. Article 14. High-risk AI systems must be designed to be effectively overseen by natural persons, who are enabled to understand, monitor, interpret, override and stop the system.
  44. Article 14. No action or decision may be taken on an identification by an Annex III point 1(a) system unless separately verified by two competent persons; not required in law enforcement, migration, border control or asylum where the law considers it disproportionate.
  45. Article 15. High-risk AI systems must achieve an appropriate and consistent level of accuracy, robustness and cybersecurity, declare their accuracy metrics, be resilient to errors and address feedback loops.
  46. Article 15. High-risk AI systems must be resilient against attempts to alter their use, outputs or performance, with measures against data poisoning, model poisoning, adversarial examples and confidentiality attacks.
  47. Article 16. Providers must indicate their name, registered trade name or mark and contact address on the system, its packaging or documentation.
  48. Article 16. Providers must have a quality management system complying with Article 17.
  49. Article 16. Providers must keep the documentation referred to in Article 18.
  50. Article 16. Providers must keep the logs automatically generated by their systems, when under their control.
  51. Article 16. The system must undergo the relevant conformity assessment procedure before being placed on the market or put into service.
  52. Article 16. Providers must draw up an EU declaration of conformity in accordance with Article 47.
  53. Article 16. Providers must affix the CE marking to the system, or its packaging or documentation.
  54. Article 16. Providers must comply with the registration obligations of Article 49(1).
  55. Article 16. Providers must take the necessary corrective actions and provide information as required in Article 20.
  56. Article 16. Upon a reasoned request of a national competent authority, providers must demonstrate the conformity of the system.
  57. Article 16. Providers must ensure the system complies with accessibility requirements under Directives (EU) 2016/2102 and (EU) 2019/882.
  58. Article 17. The quality management system must be documented as written policies, procedures and instructions covering at least the aspects in Article 17(1), proportionate to the organisation; for financial institutions, compliance with their internal governance rules stands in for most of it.
  59. Article 18. For 10 years after placing on the market or putting into service, the provider must keep the technical documentation, quality management documentation, notified body changes and decisions, and the EU declaration of conformity at the disposal of national competent authorities.
  60. Article 19. Providers must keep the automatically generated logs under their control for a period appropriate to the intended purpose, of at least six months, unless Union or national law provides otherwise.
  61. Article 20. Where the system presents a risk under Article 79(1), the provider must immediately investigate the causes and inform the market surveillance authorities and, where applicable, the notified body.
  62. Article 21. Upon a reasoned request, providers must give the competent authority all information and documentation necessary to demonstrate conformity, in an official language it understands.
  63. Article 21. Upon a reasoned request, providers must give the competent authority access to the logs under their control.
  64. Article 22. Before making a high-risk AI system available on the Union market, a provider established in a third country must appoint, by written mandate, an authorised representative established in the Union.
  65. Article 22. The provider must enable its authorised representative to perform the tasks specified in the mandate.
  66. Article 22. The authorised representative must perform the mandate: verify the declaration, documentation and conformity assessment; keep documents for 10 years; provide information; cooperate; and handle registration.
  67. Article 22. An authorised representative that considers the provider to be acting contrary to its obligations must terminate the mandate and immediately inform the market surveillance authority and the notified body.
  68. Article 23. Before placing a high-risk AI system on the market, the importer must verify the conformity assessment, technical documentation, CE marking, EU declaration, instructions for use and authorised representative.
  69. Article 23. An importer with sufficient reason to consider the system non-conforming, falsified or accompanied by falsified documentation must not place it on the market until it conforms.
  70. Article 23. Importers must indicate their name, registered trade name or mark and contact address on the system and its packaging or documentation.
  71. Article 23. While a system is under their responsibility, importers must ensure storage or transport conditions do not jeopardise its compliance.
  72. Article 23. Importers must keep the notified body certificate, instructions for use and EU declaration of conformity for 10 years.
  73. Article 23. Upon a reasoned request, importers must provide the information needed to demonstrate conformity and cooperate with the authorities in any action they take.
  74. Article 24. Before making a high-risk AI system available, distributors must verify the CE marking, EU declaration of conformity and instructions for use, and that provider and importer have complied with their obligations.
  75. Article 24. A distributor with reason to consider a system non-conforming must not make it available until it conforms, and must inform the provider or importer where it presents a risk.
  76. Article 24. While a system is under their responsibility, distributors must ensure storage or transport conditions do not jeopardise its compliance.
  77. Article 24. A distributor that considers a system it made available non-conforming must take corrective action, withdraw or recall it, or ensure the provider, importer or operator does, and inform the authorities where it presents a risk.
  78. Article 24. Upon a reasoned request, distributors must provide all information and documentation on their actions and cooperate with the authorities.
  79. Article 25. A distributor, importer, deployer or third party becomes the provider of a high-risk AI system when it puts its name on it, substantially modifies it, or modifies its intended purpose so it becomes high-risk.
  80. Article 25. Where another party becomes the provider, the initial provider must cooperate closely and make available the necessary information and access, unless it clearly specified the system is not to be changed into a high-risk system.
  81. Article 25. For a high-risk safety component of an Annex I Section A product, the product manufacturer is the provider when the system is placed on the market or put into service under the manufacturer's name.
  82. Article 25. The provider must have a written agreement with third parties supplying AI systems, tools, services, components or processes integrated in the high-risk system, except open-source suppliers of non-general-purpose components.
  83. Article 26. Deployers must take technical and organisational measures to use high-risk AI systems in accordance with the instructions for use.
  84. Article 26. Deployers must assign human oversight to natural persons with the necessary competence, training, authority and support.
  85. Article 26. To the extent it controls the input data, the deployer must ensure it is relevant and sufficiently representative for the intended purpose.
  86. Article 26. Deployers must monitor the operation of the system on the basis of its instructions for use; for financial institutions, compliance with their internal governance rules fulfils this.
  87. Article 26. A deployer with reason to consider that the system presents a risk, or that has identified a serious incident, must suspend its use and inform the provider or distributor and the market surveillance authority without undue delay.
  88. Article 26. Deployers must keep the logs under their control for a period appropriate to the intended purpose, of at least six months, unless Union or national law provides otherwise.
  89. Article 26. Before putting into service or using a high-risk AI system at the workplace, deployers who are employers must inform workers' representatives and the affected workers.
  90. Article 26. Deployers that are public authorities or Union bodies must comply with registration under Article 49; where the system is not registered in the EU database, they must not use it and must inform the provider or distributor.
  91. Article 26. Where applicable, deployers must use the information provided under Article 13 to carry out their data protection impact assessment.
  92. Article 26. In a targeted search, use of post-remote biometric identification requires authorisation requested ex ante or within 48 hours, except for the initial identification of a potential suspect on objective facts.
  93. Article 26. If the authorisation is rejected, the linked use must stop with immediate effect and the linked personal data must be deleted.
  94. Article 26. Post-remote biometric identification may never be used for law enforcement in an untargeted way, without any link to a criminal offence, proceeding, threat or missing-person search.
  95. Article 26. Each use must be documented in the relevant police file and made available to the market surveillance and data protection authorities on request.
  96. Article 26. Deployers must submit annual reports on their use of post-remote biometric identification to the market surveillance and national data protection authorities.
  97. Article 26. Deployers of Annex III systems that make or assist decisions about natural persons must inform them that they are subject to the use of the high-risk AI system.
  98. Article 26. Deployers must cooperate with the competent authorities in any action they take in relation to the high-risk AI system.
  99. Article 27. Before first deploying an Article 6(2) system (except critical infrastructure), public bodies, private providers of public services and deployers of credit-scoring or insurance-pricing systems must assess its impact on fundamental rights.
  100. Article 27. Once the fundamental rights impact assessment is performed, the deployer must notify the market surveillance authority of its results using the template.
  101. Article 31. A notified body must have legal personality under national law and meet the organisational, quality, resources, process, competence and integrity requirements, and take part in coordination.
  102. Article 31. A notified body, its management and assessment personnel must be independent of the provider and not involved in the design, development, marketing or use of the systems it assesses.
  103. Article 31. A notified body must have documented procedures to keep confidential the information it obtains, except where disclosure is required by law or to its notifying authority.
  104. Article 31. A notified body must take out liability insurance unless the Member State assumes liability or is directly responsible.
  105. Article 33. A notified body may subcontract only with the provider's agreement, to bodies meeting Article 31, taking full responsibility, listing them publicly and keeping the documents.
  106. Article 34. Notified bodies must verify conformity in accordance with the Article 43 procedures, avoid unnecessary burden, and submit documentation to the notifying authority on request.
  107. Article 36. A notified body that decides to cease its activities must inform the notifying authority and providers as soon as possible, and at least one year in advance where planned.
  108. Article 36. A notified body whose designation has been suspended, restricted or withdrawn must inform the providers concerned within 10 days.
  109. Article 40. A high-risk AI system or general-purpose AI model in conformity with published harmonised standards is presumed to conform to the requirements those standards cover.
  110. Article 41. A high-risk AI system or general-purpose AI model in conformity with the common specifications is presumed to conform to the requirements they cover.
  111. Article 41. Providers that do not comply with the common specifications must duly justify that they adopted technical solutions meeting the requirements to at least an equivalent level.
  112. Article 42. A high-risk AI system trained and tested on data reflecting the specific setting of its intended use is presumed to comply with Article 10(4).
  113. Article 42. A high-risk AI system certified under a published cybersecurity scheme is presumed to comply with the cybersecurity requirements of Article 15, as far as the certificate covers them.
  114. Article 43. For Annex III point 1 (biometrics) systems, the provider must follow the Annex VII procedure with a notified body where harmonised standards or common specifications are absent, unapplied or only partly applied.
  115. Article 43. Where harmonised standards are fully applied, the provider of an Annex III point 1 system may choose internal control (Annex VI) or the notified body procedure (Annex VII).
  116. Article 43. For high-risk AI systems in Annex III points 2 to 8, providers must follow the conformity assessment based on internal control (Annex VI).
  117. Article 43. For systems covered by Annex I Section A legislation, the provider must follow the conformity assessment required under that legislation, which includes the Section 2 requirements.
  118. Article 43. A substantially modified high-risk AI system must undergo a new conformity assessment, except changes pre-determined at the initial assessment for systems that continue to learn.
  119. Article 44. Certificates are valid for the period they indicate, at most five years for Annex I systems and four for Annex III, extendable after re-assessment.
  120. Article 44. A notified body that finds a system no longer meets Section 2 must suspend, withdraw or restrict the certificate, unless the provider corrects it within the deadline set.
  121. Article 45. Notified bodies must inform their notifying authority and the other notified bodies of certificates and approvals issued, refused or withdrawn and of other circumstances, keeping the information confidential.
  122. Article 46. A market surveillance authority may authorise placing on the market or putting into service by way of derogation from Article 43, for exceptional reasons, for a limited period.
  123. Article 46. In urgency for public security or an imminent threat, law enforcement or civil protection may use a system without the authorisation, provided it is requested during or after use without undue delay.
  124. Article 47. The EU declaration of conformity must be written, machine readable and signed, state that Section 2 is met, contain the Annex V information and be kept for 10 years; by it the provider assumes responsibility.
  125. Article 48. The CE marking must be affixed visibly, legibly and indelibly, digitally where the system is provided digitally, followed where applicable by the notified body number.
  126. Article 49. Before placing on the market or putting into service an Annex III high-risk AI system (other than point 2), the provider must register itself and the system in the EU database.
  127. Article 49. Before placing on the market an Annex III system the provider concluded is not high-risk under Article 6(3), it must register itself and the system in the EU database.
  128. Article 49. Before using an Annex III system (other than point 2), deployers that are public authorities or Union bodies must register, select the system and register its use in the EU database.
  129. Article 49. For Annex III points 1, 6 and 7 in law enforcement, migration, asylum and border control, registration is in a secure non-public section of the EU database with the listed information.
  130. Article 49. High-risk AI systems in Annex III point 2 (critical infrastructure) must be registered at national level.
  131. Article 50. Providers must design systems that interact directly with persons so they are informed they are interacting with AI, unless obvious, and except crime-detection systems not available for the public to report crimes.
  132. Article 50. Providers of systems generating synthetic audio, image, video or text must mark outputs in a machine-readable, detectable format, except assistive editing and crime-detection uses.
  133. Article 50. Deployers of emotion recognition or biometric categorisation systems must inform the persons exposed, except where permitted by law to detect, prevent or investigate criminal offences.
  134. Article 50. Deployers of systems generating or manipulating deep-fake image, audio or video content must disclose that it is artificially generated or manipulated, except where authorised by law for criminal offences.
  135. Article 50. Deployers publishing AI-generated text to inform the public on matters of public interest must disclose it, unless authorised by law for criminal offences or the text had human review with editorial responsibility.
  136. Article 50. The Article 50 information must be given clearly and distinguishably at the latest at the first interaction or exposure, meeting accessibility requirements.
  137. Article 51. A general-purpose AI model with high-impact capabilities, evaluated on the basis of appropriate technical tools and methodologies, is classified as a model with systemic risk.
  138. Article 51. A general-purpose AI model the Commission decides, ex officio or after a qualified alert, has capabilities or impact equivalent to high-impact capabilities is classified as a model with systemic risk.
  139. Article 51. A general-purpose AI model is presumed to have high-impact capabilities when its cumulative training computation exceeds 10^25 floating-point operations.
  140. Article 52. A provider whose model meets the high-impact capabilities condition may argue that, exceptionally, it does not present systemic risks; where the Commission accepts those arguments, the model is not classified as having systemic risk under Article 51(1)(a). A Commission designation under Article 51(1)(b) is unaffected.
  141. Article 52. A provider whose general-purpose AI model meets the high-impact capabilities condition must notify the Commission without delay and within two weeks of the requirement being met or becoming known.
  142. Article 53. Providers of general-purpose AI models must draw up and keep up to date the model technical documentation (Annex XI); not required for free and open-source models unless they present systemic risk.
  143. Article 53. Providers of general-purpose AI models must make available to downstream AI system providers the information they need, with at least the Annex XII elements; not required for free and open-source models unless they present systemic risk.
  144. Article 53. Providers of general-purpose AI models must put in place a policy to comply with Union copyright law, including reservations of rights.
  145. Article 53. Providers of general-purpose AI models must draw up and make public a sufficiently detailed summary of the content used for training.
  146. Article 53. Providers of general-purpose AI models must cooperate as necessary with the Commission and national competent authorities.
  147. Article 54. Before placing a general-purpose AI model on the Union market, a provider established in a third country must appoint an authorised representative in the Union by written mandate; not required for free and open-source models unless they present systemic risk.
  148. Article 54. The authorised representative of a general-purpose AI model provider must perform the mandate: verify documentation, keep it for 10 years, provide information on request, cooperate, and be addressable by the AI Office.
  149. Article 54. The authorised representative must terminate the mandate if it considers the provider to be acting contrary to its obligations, and inform the AI Office immediately.
  150. Article 55. Providers of general-purpose AI models with systemic risk must perform state-of-the-art model evaluation, including documented adversarial testing.
  151. Article 55. Providers of general-purpose AI models with systemic risk must assess and mitigate possible systemic risks at Union level.
  152. Article 55. Providers of general-purpose AI models with systemic risk must track, document and report serious incidents and corrective measures without undue delay to the AI Office.
  153. Article 55. Providers of general-purpose AI models with systemic risk must ensure adequate cybersecurity protection for the model and its physical infrastructure.
  154. Article 59. Personal data collected for other purposes may be processed in an AI regulatory sandbox only for public-interest AI systems and only when every Article 59(1) condition is met.
  155. Article 60. Testing in real world conditions outside a sandbox requires a real-world testing plan submitted to and approved by the market surveillance authority.
  156. Article 60. The testing must be registered with a Union-wide unique identification number.
  157. Article 60. The provider conducting the testing must be established in the Union or have appointed a legal representative there.
  158. Article 60. Data collected for the testing may be transferred to third countries only with appropriate safeguards under Union law.
  159. Article 60. Testing must not last longer than necessary and in any case not longer than six months, extendable once by six months on notification.
  160. Article 60. Test subjects in vulnerable groups due to age or disability must be appropriately protected.
  161. Article 60. Deployers cooperating in the testing must be informed, given instructions and have their roles agreed.
  162. Article 60. Test subjects must have given informed consent under Article 61 (with the law enforcement alternative).
  163. Article 60. The testing must be effectively overseen and the system's outputs reversible and disregardable.
  164. Article 60. Test subjects may withdraw at any time without detriment or justification and have their personal data deleted immediately and permanently.
  165. Article 60. A serious incident identified during real-world testing must be reported under Article 73, with immediate mitigation or suspension or termination of the testing.
  166. Article 60. The market surveillance authority must be notified of the suspension or termination of the testing and its final outcomes.
  167. Article 61. Freely given informed consent must follow concise, clear information on the testing, the subject's rights and the testing identification number, and be dated, documented and copied to the subject.
  168. Article 72. Providers of high-risk AI systems must establish and document a proportionate post-market monitoring system that actively and systematically collects and analyses performance data throughout the system's lifetime (an equivalent existing system under Annex I Section A legislation may be used).
  169. Article 72. The post-market monitoring system must be based on a post-market monitoring plan forming part of the technical documentation.
  170. Article 73. Providers of high-risk AI systems must report any serious incident immediately after establishing a causal link and no later than 15 days after becoming aware of it; an initial incomplete report may be followed by a complete one.
  171. Article 73. A widespread infringement or a serious incident disrupting critical infrastructure must be reported immediately and no later than two days after awareness.
  172. Article 73. In the event of the death of a person, the report must be made immediately and no later than 10 days after awareness.
  173. Article 73. After reporting a serious incident the provider must without delay investigate it, assess the risk and take corrective action.
  174. Article 73. The provider must cooperate with the authorities and must not alter the AI system in a way that may affect evaluation of the incident's causes before informing them.
  175. Article 79. Where a market surveillance authority requires corrective action, withdrawal or recall for a system presenting a risk, the operator must take it for all systems concerned it made available on the Union market.
  176. Article 80. Where a market surveillance authority requires corrective action, withdrawal or recall for a system the provider classified as non-high-risk that the authority found high-risk, the operator must take it for all systems concerned it made available on the Union market.
  177. Article 82. Where a market surveillance authority requires corrective action, withdrawal or recall for a compliant system that still presents a risk, the operator must take it for all systems concerned it made available on the Union market.
  178. Article 83. Where the authority finds the CE marking, EU declaration, registration, authorised representative or technical documentation missing or wrong, the provider must put an end to the non-compliance within the period set.
  179. Article 86. A person affected by a deployer's decision based on an Annex III high-risk system (other than point 2) with legal or similarly significant effects has the right to a clear and meaningful explanation of the AI system's role and the main elements of the decision, unless Union or national law provides otherwise.
  180. Article 91. The provider of a general-purpose AI model, or its representative, must supply the documentation and information the Commission requests, complete, correct and not misleading.
  181. Article 92. The provider of a general-purpose AI model, or its representative, must give the access to the model the Commission requests for an evaluation.
  182. Article 93. Where the Commission requests a provider of a general-purpose AI model to comply with Articles 53-54, mitigate systemic risks, or restrict, withdraw or recall the model, or makes its commitments binding, the provider must implement them.
  183. ANNEX VII. Under the Annex VII procedure, the provider must keep implementing and maintaining the quality management system as approved.
  184. ANNEX VII. Under the Annex VII procedure, any intended change to the approved quality management system, or to the list of AI systems it covers, must be brought to the notified body's attention.
  185. ANNEX VII. Under the Annex VII procedure, any change to the AI system that could affect compliance or its intended purpose must be assessed and approved by the notified body that issued the certificate.
  186. ANNEX VII. Under the Annex VII procedure, the provider must allow the notified body access to the premises where design, development and testing take place and share all necessary information.

Query this authority

People, software and AI agents can ask it whether an action may proceed and get AUTHORIZED, DENIED or ESCALATED with the rule that decided it: POST https://exchange.nomosprotocol.com/api/exchange/artifacts/eu-ai-act/query. SDK, MCP and REST.