Key facts

  • 1 August 2024 — the Act entered into force.
  • 2 February 2025 — bans on unacceptable-risk AI and AI-literacy duties (Article 4) apply.
  • 2 August 2025 — obligations for general-purpose AI (GPAI), governance bodies and penalties apply.
  • 2 August 2026 — transparency obligations (Article 50) apply; AI chat and deepfakes must be disclosed.
  • 2 December 2027 — high-risk obligations for Annex III systems, deferred under the Digital Omnibus.
  • 2 August 2028 — high-risk obligations for AI embedded in regulated products (Annex I).

What has already applied?

Since 2 February 2025, the prohibitions on unacceptable-risk AI and the AI-literacy duty under Article 4 have been in force. From 2 August 2025, obligations for general-purpose AI providers, the governance framework (including the AI Office) and the penalty provisions applied.

What changes on 2 August 2026?

The Article 50 transparency obligations apply. Providers and deployers must tell people when they are interacting with AI, and AI-generated or manipulated content such as deepfakes must be labelled. This is the date most organisations using generative AI should plan around. See what is the EU AI Act?

When do high-risk rules apply?

High-risk obligations are the strictest part of the Act and arrive in two waves: stand-alone systems listed in Annex III (such as recruitment or credit scoring), and AI embedded in products already regulated under Annex I. Regulation (EU) 2026/1744 confirms 2 December 2027 and 2 August 2028 respectively.

Did the Digital Omnibus change the timeline?

Yes. Regulation (EU) 2026/1744 was published on 24 July 2026 and entered into force on 27 July 2026. It confirms the amended high-risk dates while leaving the 2 August 2026 transparency date unchanged.

Which deadline should you focus on?

For most organisations, 2 August 2026 (transparency and GPAI governance) is the live date, alongside classifying any high-risk systems ahead of the 2027 deadlines. Use an AI inventory to know what you have before the clock runs down.

Frequently asked questions

When does the EU AI Act take effect?

Entered into force Aug 2024; obligations phase in. Prohibited practices & GPAI rules already apply (2025); transparency duties from 2 Aug 2026; high-risk deferred to 2027/2028 after the Digital Omnibus.

What is Article 50 of the EU AI Act?

Transparency obligations: telling people they're interacting with AI, and labelling AI-generated/manipulated content (e.g. deepfakes).

Did the Digital Omnibus delay the EU AI Act?

Regulation (EU) 2026/1744 confirmed high-risk obligations from 2 December 2027 for Annex III systems and 2 August 2028 for Annex I product-embedded systems.

What is the deadline I should care about most?

For most organisations: 2 Aug 2026 (transparency & GPAI governance) and your high-risk classification ahead of the 2027 deadlines.

Do I need a conformity assessment?

Only for high-risk AI systems before they're placed on the market or put into service.

Related pages

Sources

Last updated 27 July 2026.