Sign in

EU AI Act

EU Artificial Intelligence Act — European Parliament / Council

Also known as: AI Act, Regulation (EU) 2024/1689, EU Artificial Intelligence Act, AIA, Artificial Intelligence Act

Official site Official change log Subscribe to EU AI Act (RSS)

Current version
Regulation (EU) 2024/1689 as amended by Regulation (EU) 2026/1744
Published 8 July 2026. source
Status
In force
Next deadline
in 23 days
New prohibitions on non-consensual intimate imagery and child sexual abuse material apply; earlier generative AI systems must meet the Article 50(2) marking duty source
Who must comply
Providers placing AI systems or general-purpose AI models on the EU market, wherever they are based, deployers of AI systems in the EU, and non-EU providers and deployers whose AI output is used in the EU. source
How it is assessed
Most high-risk AI systems are assessed by the provider through internal control; biometric systems and products already under EU product law can require a notified body. source
Authority
Adopted by the European Parliament and the Council. National market surveillance authorities enforce it for AI systems; the Commission, through the AI Office, supervises general-purpose AI models. source source
Enforced by
National market surveillance authorities enforce the rules on AI systems, and the Commission fines general-purpose AI model providers. source
Last change

EU regulation classifying AI systems by risk level and imposing requirements on providers and deployers, with separate duties for general-purpose AI models.

Recent changes

  1. Medium
    European Commission publishes first harmonised standard EN 18286:2026 supporting AI Act implementation

    The European Commission has published EN 18286:2026 — Artificial intelligence - Quality management system for EU AI Act regulatory purposes — as the first harmonised standard …

    New document Action required

That is the last 90 days. Sign in to read further back.

Key dates

Add these dates to your calendar
  1. in 23 days New prohibitions on non-consensual intimate imagery and child sexual abuse material apply; earlier generative AI systems must meet the Article 50(2) marking duty
  2. in 266 days Providers of general-purpose AI models placed on the market before 2 August 2025 must comply (Article 111(3))
  3. in 388 days High-risk obligations apply to stand-alone AI systems classified high-risk under Article 6(2) and Annex III
  4. in 632 days High-risk obligations apply to AI systems classified high-risk under Article 6(1) and Annex I (product safety legislation)
  5. in 1362 days Providers and deployers of high-risk AI systems used by public authorities must comply (Article 111(2), as amended)
  6. in 1513 days AI systems in the EU large-scale IT systems of Annex X, placed on the market before 2 August 2027, must comply (Article 111(1))
  7. General date of application of the AI Act, including the Article 50 transparency duties (Article 113)
  8. Digital Omnibus on AI, Regulation (EU) 2026/1744, enters into force and applies to the AI Act's amending provisions (Articles 102-110)
  9. Digital Omnibus on AI, Regulation (EU) 2026/1744, adopted
  10. Obligations for general-purpose AI models, governance, notified-body rules and penalties apply (Article 113(b))
Show 3 earlier dates
  1. The Commission publishes the General-Purpose AI Code of Practice
  2. Prohibited AI practices and AI literacy (Chapters I and II) apply
  3. The AI Act enters into force

Amendments

  1. Regulation (EU) 2024/1689 as amended by Regulation (EU) 2026/1744 AmendmentCurrent text

    The Digital Omnibus on AI, adopted from proposal COM(2025) 836 and published in the Official Journal on 24 July 2026, delays the high-risk rules and simplifies other obligations. source

    More
    • High-risk obligations (Chapter III, Sections 1-3) now apply from 2 December 2027 for Annex III systems and from 2 August 2028 for Annex I systems, instead of 2 August 2026 and 2 August 2027.
    • Two new prohibitions from 2 December 2026: AI systems that generate non-consensual intimate images of identifiable people, and AI systems that generate child sexual abuse material.
    • AI literacy (Article 4) becomes a duty to take measures to support staff AI literacy, not to guarantee a specific level.
    • Generative AI systems already on the market before 2 August 2026 have until 2 December 2026 to meet the Article 50(2) marking duty.
    • Fines under Article 99(4) and (5) for small mid-cap companies (SMCs) are capped at the lower of the amount or the percentage (new Article 99(6a)); the Machinery Directive leaves Section A of Annex I and the Machinery Regulation (EU) 2023/1230 is added to Section B.
  2. Medical devices simplification proposal, COM(2025) 1023 Draft

    Commission proposal on medical devices whose Article 4 would move the medical device and in vitro diagnostic regulations from Section A to Section B of AI Act Annex I. source

    More
    • Not adopted; the European Parliament lists procedure 2025/0404(COD) as awaiting a committee decision.
  3. Regulation (EU) 2024/1689 Major version

    The Artificial Intelligence Act, published in the Official Journal on 12 July 2024. source

    More
    • It entered into force on 1 August 2024.

Who it applies to

The AI Act binds providers and deployers of AI systems and providers of general-purpose AI models connected to the EU market, with duties scaled to risk.

Providers
Providers placing AI systems on the market or putting them into service, or placing general-purpose AI models on the market, in the EU, wherever they are established (Article 2(1)(a)). source
Deployers
Deployers of AI systems established or located in the EU (Article 2(1)(b)). source
Outside the EU
Providers and deployers in third countries where the output of the AI system is used in the EU (Article 2(1)(c)). source
Risk tiers
Some practices are prohibited (Article 5); high-risk systems are those in Annex I products or Annex III use cases (Article 6); certain systems carry transparency duties (Article 50). source
Systemic-risk models
A general-purpose AI model is presumed to have systemic risk when its training used more than 10^25 floating point operations (Article 51(2)). source

How compliance is assessed

Most high-risk AI systems are assessed by the provider through internal control; biometric systems and products already under EU product law can require a notified body.

Annex III, points 2-8
Internal control under Annex VI, without a notified body (Article 43(2)). source
Biometrics (Annex III point 1)
Internal control is allowed only where harmonised standards or common specifications are applied; otherwise a notified body assesses under Annex VII (Article 43(1)). source
Annex I products
The conformity procedure of the relevant product legislation applies and includes the AI Act requirements (Article 43(3)). source
Re-assessment
A substantial modification requires a new conformity assessment; notified-body certificates last up to five years (Annex I) or four years (Annex III) (Articles 43(4) and 44(2)). source
Registration
Providers register Annex III high-risk systems in the EU database before placing them on the market, except critical-infrastructure systems (Article 49(1)). source
General-purpose AI
The General-Purpose AI Code of Practice is a voluntary tool providers can use to demonstrate compliance. source

Enforcement and penalties

National market surveillance authorities enforce the rules on AI systems, and the Commission fines general-purpose AI model providers.

Up to EUR 35 million or 7%

Engaging in a prohibited AI practice (Article 5); for an undertaking, 7% of total worldwide annual turnover if higher (Article 99(3)). source

Up to EUR 15 million or 3%

Breaching the obligations of providers, deployers, importers, distributors, authorised representatives or notified bodies, including transparency duties (Article 99(4)). source

Up to EUR 7.5 million or 1%

Supplying incorrect, incomplete or misleading information to notified bodies or national authorities (Article 99(5)). source

Up to 3% or EUR 15 million

Commission fines on providers of general-purpose AI models, whichever is higher (Article 101(1)). source

National penalties
Member States lay down penalties, which may include fines, warnings and non-monetary measures, within the ceilings of Article 99 (Article 99(1), as amended). source
SMEs and small mid-caps
For SMEs and start-ups each fine is capped at the lower of the amount or the percentage (Article 99(6)); for small mid-cap companies the same applies to fines under Article 99(4) and (5) (new Article 99(6a)). source source

Documents

  1. Law text Regulation (EU) 2024/1689 (Artificial Intelligence Act)
  2. Law text Regulation (EU) 2026/1744 (Digital Omnibus on AI)
  3. Guidance The General-Purpose AI Code of Practice
  4. FAQ General-purpose AI models in the AI Act - questions and answers
  5. Law text Medical devices simplification proposal COM(2025) 1023 (amends AI Act Annex I; not adopted)

Verified against official sources on 18 October 2026. We monitor the official sources for this framework. Watching since 7 September 2026. Subscribe to EU AI Act (RSS)