EU AI Policy CheckerPreliminary EU AI Act assessment

Sources

Every rule, and where it comes from

All 53 rules in the current set, read from the Official Journal text of Regulation (EU) 2024/1689 published on 12 July 2024. The page numbers match that document, so any line here can be checked against it.

Every rule is marked as requiring qualified review. None has been signed off by a named reviewer. Where a rule rests on an interpretation rather than a mechanical reading, the result screen says so on the finding itself.

Rule set 2024-1689.2026-09-17.1. Sorted by page in the Official Journal.
ProvisionPageWhat it decidesApplies fromReview
Article 2(1)45Article 2(1) reaches your situation, so the Regulation applies to you on territorial grounds.2026-08-02Needs review
Article 2(3)45Article 2(3) takes this outside the Regulation, because the system is used exclusively for military, defence or national security purposes.2026-08-02Needs review
Article 2(4)45Article 2(4) may take this outside the Regulation, because a third-country public authority or international organisation is using the system under a law enforcement or judicial cooperation framework with the Union.2026-08-02Needs review
Article 2(2)45Article 2(2) limits which parts of the Regulation apply, because this is a high-risk system relating to a product covered by Annex I Section B legislation.2026-08-02Needs review
Article 3(1)46What you described meets the definition of an AI system in Article 3(1).2025-02-02Needs review
Article 2(6)46Article 2(6) takes this outside the Regulation, because the system or model was developed and put into service for the sole purpose of scientific research and development.2026-08-02Needs review
Article 2(8)46Article 2(8) takes this outside the Regulation for now, because the activity is research, testing or development before the system is placed on the market or put into service.2026-08-02Needs review
Article 2(10)46Article 2(10) removes deployer obligations here, because a natural person is using the system in a purely personal, non-professional activity.2026-08-02Needs review
Article 2(12)46Article 2(12) takes this outside the Regulation, because the AI system is released under a free and open-source licence and none of the carve-backs apply.2026-08-02Needs review
Article 3(63)50What you described meets the definition of a general-purpose AI model in Article 3(63).2025-08-02Needs review
Article 5(1)(a)51Article 5(1)(a) may be engaged. The system appears to use subliminal, purposefully manipulative or deceptive techniques that materially distort behaviour and cause or are likely to cause significant harm.2025-02-02Needs review
Article 5(1)(b)51Article 5(1)(b) may be engaged. The system appears to exploit vulnerabilities due to age, disability or a specific social or economic situation in a way that materially distorts behaviour and causes or is likely to cause significant harm.2025-02-02Needs review
Article 5(1)(c)51Article 5(1)(c) may be engaged. The system appears to score people over time on social behaviour or personal characteristics, with detrimental treatment following.2025-02-02Needs review
Article 5(1)(d)51Article 5(1)(d) may be engaged. The system appears to predict the risk of a person committing a criminal offence based solely on profiling or on assessing personality traits.2025-02-02Needs review
Article 5(1)(e)51Article 5(1)(e) may be engaged. The system appears to create or expand facial recognition databases through untargeted scraping of facial images from the internet or CCTV footage.2025-02-02Needs review
Article 5(1)(f)51Article 5(1)(f) may be engaged. The system appears to infer emotions of people in a workplace or an education institution.2025-02-02Needs review
Article 451Article 4 applies. Providers and deployers have to take measures to ensure a sufficient level of AI literacy among staff and others operating the system on their behalf.2025-02-02Needs review
Article 5(1)(g)52Article 5(1)(g) may be engaged. The system appears to categorise people individually from their biometric data to deduce or infer race, political opinions, trade union membership, religious or philosophical beliefs, sex life or sexual orientation.2025-02-02Needs review
Article 5(1)(h)52Article 5(1)(h) may be engaged. This is real-time remote biometric identification in a publicly accessible space for law enforcement purposes.2025-02-02Needs review
Article 6(1) Annex I53Article 6(1) is met. The AI system is a product, or a safety component of a product, covered by Annex I legislation, and that product needs a third-party conformity assessment.2027-08-02
reported 2028-08-02
Needs review
Article 6(2) Annex III53Article 6(2) points to a high-risk classification, because the system matches an Annex III use case and the Article 6(3) derogation does not carry it out.2026-08-02
reported 2027-12-02
Needs review
Article 6(4)54Article 6(4) applies. A provider relying on the Article 6(3) derogation has to document that assessment before placing the system on the market, and register under Article 49(2).2026-08-02
reported 2027-12-02
Needs review
Article 1662Article 16 applies. As provider of a high-risk AI system you carry the full set of duties in points (a) to (l).2026-08-02
reported 2027-12-02
Needs review
Article 2667Article 26 applies. As deployer of a high-risk AI system you carry the operational duties in paragraphs 1 to 11.2026-08-02
reported 2027-12-02
Needs review
Article 25(1)67Article 25(1) applies. Your organisation takes on provider duties under Article 16 for this high-risk system, even though it started as a distributor, importer, deployer or third party.2026-08-02
reported 2027-12-02
Needs review
Article 26(7)68Article 26(7) applies. Before putting a high-risk system into service at the workplace, an employer has to inform workers representatives and the affected workers.2026-08-02
reported 2027-12-02
Needs review
Article 27(1)69Article 27 applies. A fundamental rights impact assessment is needed before the first use of this high-risk system.2026-08-02
reported 2027-12-02
Needs review
Article 50(1)82Article 50(1) applies. People interacting with the system have to be told they are interacting with an AI system.2026-08-02Needs review
Article 50(2)82Article 50(2) applies. Outputs have to be marked in a machine-readable format and be detectable as artificially generated or manipulated.2026-08-02
reported 2026-12-02
Needs review
Article 50(3)82Article 50(3) applies. People exposed to the system have to be informed that it is operating.2026-08-02Needs review
Article 50(4)82Article 50(4) first subparagraph applies. The deployer has to disclose that the content has been artificially generated or manipulated.2026-08-02Needs review
Article 50(4)82Article 50(4) second subparagraph applies. Text published to inform the public on matters of public interest has to be disclosed as artificially generated or manipulated.2026-08-02Needs review
Article 50(5)83Article 50(5) applies. Any disclosure owed under Article 50 has to be clear, distinguishable, given no later than the first interaction or exposure, and accessible.2026-08-02Needs review
Article 51(1 and 2)83Article 51 points to a general-purpose AI model with systemic risk, either on the Article 51(2) compute presumption or by Commission decision.2025-08-02Needs review
Article 53(1)84Article 53(1) applies. As the provider of a general-purpose AI model you owe the four duties in points (a) to (d).2025-08-02Needs review
Article 53(2)85Article 53(2) relieves you of the documentation duties in Article 53(1)(a) and (b), because the model is released under a qualifying free and open-source licence.2025-08-02Needs review
Article 54(1)85Article 54(1) applies. A provider established in a third country has to appoint an authorised representative in the Union by written mandate before placing the model on the Union market.2025-08-02Needs review
Article 55(1)86Article 55(1) applies. On top of Articles 53 and 54 you owe model evaluation, systemic risk assessment and mitigation, serious incident reporting, and cybersecurity protection.2025-08-02Needs review
Article 111(2)121Article 111(2) may change your deadline. High-risk systems placed on the market or put into service before 2 August 2026 are reached only if they undergo significant changes in design after that date.2026-08-02Needs review
Article 111(3)121Article 111(3) applies. Providers of general-purpose AI models placed on the market before 2 August 2025 have until 2 August 2027 to comply.2025-08-02Needs review
Article 113(third subparagraph)(a)123On the date you gave, Chapters I and II are in application, so the AI literacy duty and every Article 5 prohibition already bite.2025-02-02Needs review
Article 113(third subparagraph)(b)123On the date you gave, Chapter V is in application, so the general-purpose AI model duties already bite.2025-08-02Needs review
Article 113(second subparagraph)123On the date you gave, the general application date of the Regulation has passed, so Article 50 and the rest of the Regulation apply unless another date is carved out.2026-08-02Needs review
Article 113(second subparagraph)123On the date you gave, the Annex III high-risk regime is in application on the face of the Regulation as published.2026-08-02
reported 2027-12-02
Needs review
Article 113(third subparagraph)(c)123Article 6(1) and its corresponding obligations apply from 2 August 2027 on the face of the Regulation as published.2027-08-02
reported 2028-08-02
Needs review
Article 6(2) Annex III point 1127What you described matches Annex III point 1 on biometrics.2026-08-02
reported 2027-12-02
Needs review
Article 6(2) Annex III point 2127What you described matches Annex III point 2 on critical infrastructure.2026-08-02
reported 2027-12-02
Needs review
Article 6(2) Annex III point 3127What you described matches Annex III point 3 on education and vocational training.2026-08-02
reported 2027-12-02
Needs review
Article 6(2) Annex III point 4127What you described matches Annex III point 4 on employment, workers management and access to self-employment.2026-08-02
reported 2027-12-02
Needs review
Article 6(2) Annex III point 5127What you described matches Annex III point 5 on access to essential private and public services.2026-08-02
reported 2027-12-02
Needs review
Article 6(2) Annex III point 6128What you described matches Annex III point 6 on law enforcement.2026-08-02
reported 2027-12-02
Needs review
Article 6(2) Annex III point 7128What you described matches Annex III point 7 on migration, asylum and border control management.2026-08-02
reported 2027-12-02
Needs review
Article 6(2) Annex III point 8128What you described matches Annex III point 8 on the administration of justice and democratic processes.2026-08-02
reported 2027-12-02
Needs review

The source document

Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 laying down harmonised rules on artificial intelligence. Published in the Official Journal on 12 July 2024. The European Legislation Identifier is data.europa.eu/eli/reg/2024/1689/oj.

That text is the version as published. It cannot show amendments made afterwards. Where a change has been reported but not verified against primary text, the table above marks it and the result screen explains it.