Appearance
EU AI Act
RegulationEuropean Union
Harmonized rules on artificial intelligence
The EU AI Act is Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024. It gives harmonized rules on artificial intelligence. Regulation (EU) 2026/1744, the Digital Omnibus on AI, amended it in 2026. This page gives the scope, which obligations apply to whom and from when, and the dates. The other pages of this section give each obligation with its conditions, the penalties and enforcement, and the Commission guidance.
Key facts
Legislator
European Parliament and Council
Instrument
Regulation (EU) 2024/1689 of 13 June 2024, amended in 2026 by Regulation (EU) 2026/1744, the Digital Omnibus on AI
In force
1 August 2024. The Digital Omnibus on AI entered into force on 27 July 2026.
Applies
General date of application 2 August 2026. AI literacy and most prohibitions from 2 February 2025. General-purpose AI models from 2 August 2025. The two new prohibitions from 2 December 2026. Most high-risk rules from 2 December 2027 (Annex III) or 2 August 2028 (Annex I). Article 113, as amended. Article 111 gives transitional rules for some AI systems and general-purpose AI models that were placed on the market or put into service before the dates in that Article (see Dates).
Covers
Providers, deployers, and other actors in relation to AI systems and general-purpose AI models in the Union
Enforced by
National competent authorities, the European Data Protection Supervisor (EDPS), and the Commission, which acts through the AI Office
Maximum fines
For an undertaking, EUR 35 000 000 or 7% of total worldwide annual turnover for the preceding financial year, whichever is higher, for non-compliance with the Article 5 prohibitions (Article 99(3)). For SMEs, the lower of the two (Article 99(6)). Lower maximums apply to other infringements.
Modulos Frameworks
MFF-1 (AI System) and OFF-1 (Organization)
Which obligations apply and from when
Different obligations of the EU AI Act apply from different dates (Article 113, as amended by Regulation (EU) 2026/1744). Find the row that applies to you:
| Your role or AI system | Obligations apply from | Read |
|---|---|---|
| Provider or deployer of an AI system (AI literacy, Article 4) | 2 February 2025 | Article 4 — AI literacy |
| AI system, or its use, that is a prohibited AI practice in Article 5 | 2 February 2025. For Article 5(1), points (ba) and (bb), and Article 5(1a) and (1b): 2 December 2026 | Article 5 — prohibited AI practices |
| Provider of a general-purpose AI model (Chapter V) | 2 August 2025. For models placed on the market before 2 August 2025: by 2 August 2027 (Article 111(3)). The Commission powers in Articles 91 to 93 and the fines in Article 101 apply from 2 August 2026 | General-purpose AI models |
| Provider or deployer of an AI system that Article 50 covers (transparency obligations) | 2 August 2026. For providers of AI systems that generate synthetic audio, image, video, or text content and that were placed on the market before 2 August 2026: Article 50(2) by 2 December 2026 (Article 111(4)) | Article 50 — transparency obligations |
| Provider, deployer, or other operator of a high-risk AI system in Article 6(2) and Annex III | 2 December 2027, for Chapter III Sections 1, 2, and 3, but not Article 6(5) | High-risk AI systems |
| Provider, deployer, or other operator of a high-risk AI system in Article 6(1) and Annex I | 2 August 2028, for Chapter III Sections 1, 2, and 3, but not Article 6(5). For Section B of Annex I, Article 2(2) limits the provisions that apply | High-risk AI systems and Article 2(2) |
| High-risk AI system placed on the market or put into service before its Chapter III date | The EU AI Act applies to its operators only if, from that date, the system is subject to "significant changes in their designs" (Article 111(2)). Providers and deployers of high-risk AI systems "intended to be used by public authorities" must "take the necessary steps to comply" by 2 August 2030 | Dates |
Some provisions for high-risk AI systems apply from the general date, 2 August 2026. Examples are Chapter III Section 5 (standards, conformity assessment, certificates, registration) and Articles 72 and 73 (post-market monitoring and serious incidents). Article 111(1) gives a different rule for AI systems that are components of the large-scale IT systems in Annex X. Article 111(1) and (2) apply "without prejudice to the application of Article 5". For all dates and transitional rules, see Dates.
EU AI Act scope
The EU AI Act gives rules for providers, deployers, and other actors in relation to AI systems and general-purpose AI models in the Union. Article 1(1) gives its purpose:
The purpose of this Regulation is to improve the functioning of the internal market and promote the uptake of human-centric and trustworthy artificial intelligence (AI), while ensuring a high level of protection of health, safety, fundamental rights enshrined in the Charter, including democracy, the rule of law and environmental protection, against the harmful effects of AI systems in the Union and supporting innovation.
The EU AI Act is a Regulation. It is "binding in its entirety and directly applicable in all Member States" (Article 113).
The Act has two legal bases in the Treaty on the Functioning of the European Union (TFEU): Article 114 for the internal market and Article 16 for data protection. Article 16 supports "specific rules on the processing of personal data" for prohibitions concerning remote biometric identification, biometric categorization, and individual risk assessments for law enforcement. See the Commission prohibited-practices guidelines, paragraph 10.
The Article 2 scope tests and exclusions give the conditions for EU AI Act provisions to apply to providers, deployers, and other actors. Article 2(2) limits the provisions applicable to Article 6(1) high-risk systems related to Annex I Section B products.
Article 3 defines AI systems, intended purpose, and actor roles.
Digital Omnibus on AI
The Digital Omnibus on AI amended the EU AI Act. This amendment is Regulation (EU) 2026/1744 of 8 July 2026, which also amends Regulations (EU) 2018/1139 and (EU) 2023/1230. It is in the Official Journal of 24 July 2026. It entered into force on 27 July 2026, the third day after its publication (Article 4 of Regulation (EU) 2026/1744). The consolidated text of the EU AI Act of 27 July 2026 contains the amendments.
These are some of the changes that Regulation (EU) 2026/1744 made to the EU AI Act:
- Application dates — Articles 113 and 111 give the amended application dates and transitional rules.
- Two new prohibitions — Article 5 adds conditional prohibitions concerning intimate material without consent and child sexual abuse material.
- Rules for safety components — Article 6(1a) to (1c) give rules for safety components and the third-party conformity assessment condition in Article 6(1)(b).
- Machinery — The amendment moved machinery legislation to Annex I Section B, with the Article 2(2) scope limit for related Article 6(1) high-risk systems.
- Limits for Section A of Annex I — Article 2(13) allows conditional limits to specified requirements and obligations for Article 6(1) high-risk systems related to Section A products. The Commission must adopt delegated acts that specify these limits by 2 August 2027.
- Bias detection and correction — Article 4a gives specified providers and deployers a conditional permission to process special categories of personal data for bias detection and correction.
- SMEs and SMCs — Articles 11(1) and 17(2) address simplified technical documentation and proportionate quality management for providers, with specified rules for SMEs and SMCs.
- AI Office — Article 75 gives the AI Office supervision and enforcement competences for specified AI systems.
Dates
The EU AI Act has different application dates for different provisions. The EU AI Act entered into force on 1 August 2024 (Article 113). This is the twentieth day after its publication in the Official Journal of 12 July 2024. The general date of application is 2 August 2026. These are the dates of Article 113, as amended by Regulation (EU) 2026/1744:
| Date | Provisions that apply | Provision of Article 113 |
|---|---|---|
| From 2 February 2025 | Chapters I and II, for example the definitions, Article 4 on AI literacy, and Article 5(1), points (a) to (h). Not Article 5(1), points (ba) and (bb), and not Article 5(1a) and (1b) | Point (a) |
| From 2 August 2025 | Chapter III Section 4 (notifying authorities and notified bodies), Chapter V (general-purpose AI models), Chapter VII (governance), Chapter XII (penalties), and Article 78. Not Article 101 | Point (b) |
| From 27 July 2026 | Articles 102 to 110, which amend other Union acts | Point (d) |
| From 2 August 2026 | All provisions that the other rows do not name, for example Chapter III Section 5, Article 50, Chapter IX, and Article 101 | Second paragraph |
| From 2 December 2026 | Article 5(1), points (ba) and (bb), and Article 5(1a) and (1b) | Point (a) |
| From 2 December 2027 | Chapter III Sections 1, 2, and 3, but not Article 6(5), for AI systems classified as high-risk in Article 6(2) and Annex III | Point (c)(i) |
| From 2 August 2028 | Chapter III Sections 1, 2, and 3, but not Article 6(5), for AI systems classified as high-risk in Article 6(1) and Annex I. For Section B of Annex I, Article 2(2) also applies | Point (c)(ii) |
Chapter III Sections 1, 2, and 3 contain Articles 6 to 27. They give the classification rules, the requirements for high-risk AI systems, and the obligations of providers, deployers, and other parties.
Article 111 gives transitional rules for AI systems and general-purpose AI models that were placed on the market or put into service before these dates:
- Article 111(1) — This rule is for AI systems that are components of the large-scale IT systems in Annex X. It applies if these AI systems were placed on the market or put into service before 2 August 2027. They must be brought into compliance with the EU AI Act by 31 December 2030. This rule applies "without prejudice to the application of Article 5".
- Article 111(2) — This rule is for other high-risk AI systems placed on the market or put into service before their Chapter III date. The EU AI Act applies to the operators of these systems only if, from that date, the systems are "subject to significant changes in their designs". In all cases, providers and deployers of high-risk AI systems "intended to be used by public authorities" must "take the necessary steps to comply" by 2 August 2030. This rule also applies "without prejudice to the application of Article 5".
- Article 111(3) — Providers of general-purpose AI models placed on the market before 2 August 2025 must "take the necessary steps in order to comply" by 2 August 2027.
- Article 111(4) — This rule is for providers of AI systems, also general-purpose AI systems, that generate synthetic audio, image, video, or text content. If these systems were placed on the market before 2 August 2026, the providers must "take the necessary steps in order to comply" with Article 50(2) by 2 December 2026.
Recital 39 of Regulation (EU) 2026/1744 clarifies that the Article 111(2) grace period applies to a type and model of high-risk AI system. This applies if the first unit was lawfully placed on the Union market or put into service before the Article 111(2) date. Subsequent units of that same type and model can also benefit while "the design of that high-risk AI system remains unchanged". They can continue to be placed on the market, made available, or put into service "without any additional obligations, requirements or the need for additional certification". Article 5 and the public-authority deadline above continue to apply.
The test of Article 111(2), "significant changes in their designs", is different from the "substantial modification" of Article 3(23).
Article 57(1) tells each Member State to make sure that its competent authorities "establish at least one AI regulatory sandbox at national level". That sandbox "shall be operational by 2 August 2027". It "may also be established jointly with the competent authorities of other Member States". The obligation "may also be fulfilled by participating in an existing sandbox" that gives an equivalent level of national coverage. Article 57(3a) permits the AI Office to "establish an AI regulatory sandbox at Union level for AI systems covered by Article 75(1)".
Find the applicable obligations
Each item below gives one group of EU AI Act rules in short, with a link to the page that gives its conditions and exceptions.
For Article 6(1) high-risk systems related to Annex I Section B products, only Article 6(1), Article 60a, and Articles 102 to 112 apply. Articles 57, 58, and 59 apply to them only "in so far as the requirements for high-risk AI systems under this Regulation have been integrated in that Union harmonisation legislation" (Article 2(2)).
- AI literacy (Article 4) — In-scope providers and deployers must "take measures to support the development of AI literacy" as specified in Article 4, without a duty to "guarantee any specific level of AI literacy of any individual".
- Prohibited AI practices (Article 5) — Article 5 prohibits specified AI practices, with different acts, conditions, and exceptions for each prohibition.
- Transparency (Article 50) — Article 50 gives providers and deployers different transparency duties, with specified triggers and exceptions, for high-risk and other AI systems.
- High-risk AI systems (Article 6) — Article 6 gives the cumulative product-related tests, the Annex III classification rules, and the conditional Article 6(3) exception.
- General-purpose AI models (Chapter V) — Chapter V gives duties to model providers, with more duties for systemic-risk models. Model-provider and system-provider duties are different and can apply to the same organization.
- Conformity assessment (Article 43) — Providers must make sure in-scope high-risk AI systems undergo the applicable Article 43 assessment before placing on the market or putting into service, subject to Article 46.
- Post-market monitoring and serious incidents (Articles 72 and 73) — Providers of in-scope high-risk AI systems must comply with the applicable monitoring and serious-incident reporting duties of Articles 72 and 73.
Penalties and enforcement
National competent authorities, the European Data Protection Supervisor (EDPS), and the Commission, which acts through the AI Office, enforce the EU AI Act. Articles 99 to 101 give these maximum fines:
| Infringement | Maximum fine | Provision |
|---|---|---|
| Non-compliance with the prohibition of the AI practices in Article 5 | EUR 35 000 000 or 7% | Article 99(3) |
| Non-compliance with the provisions that Article 99(4) lists, for example Articles 16, 26, and 50 | EUR 15 000 000 or 3% | Article 99(4) |
| The supply of "incorrect, incomplete or misleading information to notified bodies or national competent authorities in reply to a request" | EUR 7 500 000 or 1% | Article 99(5) |
| Non-compliance by Union institutions, bodies, offices, and agencies (fines by the EDPS) | EUR 1 500 000 for Article 5. EUR 750 000 for other requirements or obligations | Article 100(2) and (3) |
| Providers of general-purpose AI models (fines by the Commission) | 3% or EUR 15 000 000, "whichever is higher" | Article 101(1) |
For an undertaking, the percentage is of its total worldwide annual turnover for the preceding financial year, and the higher of the two values is the maximum. For SMEs, also start-ups, the lower value is the maximum for each fine in Article 99 (Article 99(6)). For SMCs, the lower value applies only to the fines in Article 99(4) and (5) (Article 99(6a)). Article 101 has no such rule: for providers of general-purpose AI models, the higher value is the maximum. For the authorities, their powers, the enforcement dates, and the conditions of each fine, see Penalties and enforcement.
Guidance status
The European Commission gives guidelines and other guidance on the EU AI Act. Commission guidelines are not binding, and they do not change the provisions of the EU AI Act. Article 99(1) tells the Member States to take the Commission guidelines of Article 96 into account in their rules on penalties (see Enforcement and penalties). For the status of each guidance document, adopted or draft, see Publication and adoption dates and Other Commission guidance.
Article 95 also provides for voluntary codes of conduct. Paragraph 1 concerns voluntary application of some or all Chapter III Section 2 requirements to AI systems that are not high-risk. Paragraph 2 concerns voluntary application of "specific requirements" to all AI systems, also by deployers. These codes of conduct are different from the Article 50 and Chapter V codes of practice.
Standards and codes of practice have a different status:
- EN 18286:2026 — Decision (EU) 2026/2274 published the standard's "reference" (OJ, 14 October 2026), with presumption of conformity only for Article 17(1) and the first sentence of Article 11(1), subject to the conditions in each row (Annex ZA, Table ZA.1).
- Code of Practice on Transparency of AI-generated Content — This voluntary tool for providers and deployers supports compliance with the EU AI Act.
- General-Purpose AI Code of Practice — Until a harmonized standard is published, providers of general-purpose AI models can use codes of practice to demonstrate compliance (Articles 53(4) and 55(2)). The Commission and the AI Board found this code to be an "adequate voluntary tool" on 1 August 2025. See Article 56 — Codes of Practice.
The EU AI Act in Modulos
Modulos provides EU AI Act Frameworks OFF-1 at Organization level and MFF-1 at AI System level, with setup decisions and the Requirement table. For the product page, see EU AI Act compliance on modulos.ai.
EU AI Act FAQ
For short answers about the EU AI Act, see Frequently asked questions about the EU AI Act.
Related information
For more on the EU AI Act and related instruments, see these pages:
- Commission guidance on the EU AI Act
- Harmonized standards for the EU AI Act
- EU AI Act and GDPR, ISO/IEC 42001 and EU AI Act, and NIST AI RMF and EU AI Act
- DORA and the EU AI Act
- Control overlap between the Modulos Frameworks
Sources
- EU AI Act, Regulation (EU) 2024/1689 of 13 June 2024, OJ L, 12.7.2024 (EUR-Lex). Consolidated text of 27 July 2026 (EUR-Lex): Articles 1 to 6, 11(1), 16(f), 17(2), 43, 46, 50, 51, 53(4), and 55(2). Also Articles 57 to 59, 60a, 72, 73, and 75. Also Articles 91 to 93, 95(1) and (2), 99 to 113, and Annex I.
- Regulation (EU) 2026/1744 of 8 July 2026, the Digital Omnibus on AI, OJ L, 24.7.2026: recital 39 and Article 4 (EUR-Lex).
- European Commission, Guidelines on prohibited artificial intelligence practices, "C(2025) 5052 final" of 29 July 2025: paragraph 10.
- Commission Implementing Decision (EU) 2026/2274 of 12 October 2026, OJ L, 14.10.2026: Articles 1 and 2 and the Annex (EUR-Lex).
- European Commission, “Commission opinion on the assessment of the Code of Practice on Transparency of AI-generated Content”, 9 July 2026.
- European Commission, “Commission Opinion on the assessment of the General-Purpose AI Code of Practice”, 1 August 2025.
- European Commission, “AI Act”, last update 3 August 2026: the date of entry into force.
Disclaimer
This page gives general information. It is not legal advice.