Skip to content
Amended. Regulation (EU) 2026/1744 entered into force 27 July 2026. See what moved →
EU AI Act ChecklistIndependent reference
Application dates · Article 113

The EU AI Act timeline, as amended

Eight dates, and the provisions that attach to each. This page reflects Regulation (EU) 2024/1689 as amended by Regulation (EU) 2026/1744, which entered into force on 27 July 2026 and fixed the high-risk dates as calendar dates rather than standards-conditional triggers.

Art. 113Reg. (EU) 2026/1744in force 27 Jul 2026

The eight dates

Dates in the past are marked as applying. Nothing below is conditional on the availability of harmonised standards: the adopted Omnibus text removed that mechanism.

  1. 2 Feb 2025 applies

    Article 5 prohibited practices (original list). Article 4 AI literacy.

  2. 2 Aug 2025 applies

    General-purpose AI model obligations, Chapter V (Arts. 51–55).

  3. 2 Aug 2026 applies

    Article 50 transparency. Article 49 registration. National market surveillance authority powers. General application date.

  4. 2 Dec 2026 next

    Two new Article 5 prohibitions (NCII/NCIM and CSAM generation). Article 50(2) machine-readable marking for generative systems already on the market before 2 Aug 2026.

  5. 2 Aug 2027 upcoming

    National regulatory sandboxes (deferred by Reg. 2026/1744).

  6. 2 Dec 2027 upcoming

    Annex III stand-alone high-risk obligations — Chapter III Sections 1, 2 and 3.

  7. 2 Aug 2028 upcoming

    Annex I embedded high-risk obligations (AI in regulated products).

  8. 2 Aug 2030 upcoming

    Article 111 transition ends for AI systems already in use by public authorities.

What Reg. (EU) 2026/1744 moved
Annex III stand-alone high-risk 2 Aug 20262 Dec 2027
Annex I embedded high-risk 2 Aug 20272 Aug 2028
Art. 50(2) marking, systems already on market 2 Aug 20262 Dec 2026
National regulatory sandboxes 2 Aug 20262 Aug 2027
New Art. 5 prohibitions (NCII/NCIM, CSAM) 2 Dec 2026
Art. 5 (original) · Ch. V GPAI · Art. 49 · Art. 50 (most) unchangedunchanged
Reg. (EU) 2026/1744OJ 24 Jul 2026in force 27 Jul 2026

Where this usually goes next

Three situations account for most people reading this page. Each has a different answer.

A deal is blocked on an AI questionnaire

Legal will not sign until you can evidence how AI is governed. HumanAudit’s AI Trust Package is a fixed $3,500 over five business days: a public trust page, a pre-filled SIG Lite / CAIQ / SSPA Section K questionnaire bank, and your AI inventory and classification.

How this works for AI companies →

You need ISO/IEC 42001 documentation

23 clause-mapped AIMS documents with all 38 Annex A controls pre-populated, editable and yours to keep, from $199. Or score your gaps first: 18 questions, free, no signup to begin.

Free gap assessment →
See the three tiers →

You are not sure what reaches you

Twenty minutes with the founder. No prep, no deck, straight to the person accountable for the work. If none of this applies to you, you get told that on the call.

Book a free 20-minute call →

This reference is published by HumanAudit Inc. Not a law firm, not an accredited certification body, not a registered auditor. We build documentation, your counsel interprets it, and an accredited body of your choosing certifies you. How this is funded →

What the deferral actually covers

The deferral concerns Chapter III Sections 1, 2 and 3 only. That is: classification rules for high-risk systems, the Section 2 requirements (Articles 8–15), and the Section 3 obligations on providers and other operators (Articles 16–27).

It does not cover:

  • Article 5 prohibitions — in force since 2 February 2025, and now longer, not shorter.
  • Chapter V general-purpose AI models, Articles 51–55 — in force since 2 August 2025, untouched in substance.
  • Article 50 transparency — 2 August 2026, with only the narrow marking transition moved.
  • Article 49 registration — 2 August 2026.
  • Article 4 AI literacy — 2 February 2025, at every risk tier.
  • Penalty ceilings: unchanged.

The sentence to be careful with

“The EU delayed the AI Act” and “the high-risk chapter was deferred” are different statements. The first is what most coverage said. The second is what happened. The difference is your entire live obligation set.

What this means for planning

Practical recommendation. Sixteen extra months is not sixteen months of slack, for one reason: the expensive part of high-risk compliance is not writing documents, it is finding and classifying every AI system and keeping that inventory alive as products ship. That work is independent of the standards and independent of the date.

Do nowWhy
Article 50 disclosureLive. Penalty-bearing. Usually a product change, not a document.
Article 4 literacy recordsLive since Feb 2025 and the most commonly undocumented obligation.
Synthetic content marking2 Dec 2026 for pre-existing systems. Provenance marking is engineering work with a lead time.
NCII / CSAM safeguards2 Dec 2026. If you ship image or multimodal generation, this is a design decision now.
AI system inventoryGates classification, registration, conformity assessment and every questionnaire you will be sent.

Status labels on this page

Verified fact: The eight dates and the scope of the deferral, checked against the adopted text of Reg. (EU) 2026/1744 and Tier-2 analysis of it.

Expert analysis: The planning table and the argument that the deferral does not create slack.

Unsettled: Operational readiness of national authorities and the EU database varies by member state.

Next step

Turn dates into a plan

The useful artefact is not a timeline, it is a mapping from each date to the systems it touches and the person who owns the work. That mapping starts with an inventory.

Not sure where you sit?

The classifier maps your system against Articles 5, 6, 50 and Annex III. Twelve questions, no email.

Run the classifier →

Frequently asked

What are the EU AI Act deadlines after the Digital Omnibus?

Six dates matter. 2 February 2025: Article 5 prohibitions and Article 4 AI literacy. 2 August 2025: general-purpose AI model obligations. 2 August 2026: Article 50 transparency, Article 49 registration, national market surveillance authority powers, and the general application date. 2 December 2026: two new Article 5 prohibitions covering non-consensual intimate imagery and CSAM generation, plus Article 50(2) machine-readable marking for generative systems already on the market before 2 August 2026. 2 December 2027: Annex III stand-alone high-risk obligations. 2 August 2028: Annex I embedded high-risk obligations.

Was the EU AI Act delayed?

Partly, and precisely. Regulation (EU) 2026/1744 deferred Chapter III Sections 1, 2 and 3, which cover classification, requirements and provider obligations for high-risk AI systems, from 2 August 2026 to 2 December 2027 for stand-alone Annex III systems, and from 2 August 2027 to 2 August 2028 for Annex I embedded systems. It did not defer the Article 5 prohibitions, the Chapter V general-purpose AI regime, Article 50 transparency, Article 49 registration, or the general application date of 2 August 2026.

Are the new high-risk dates conditional on standards being ready?

No. The Commission's original proposal linked the high-risk application date to a Commission decision confirming that harmonised standards were available, with backstop dates. The adopted text removed that conditional trigger and set fixed calendar dates instead.

What happened on 2 August 2026?

The general application date arrived. Article 50 transparency obligations took effect, requiring disclosure that a person is interacting with an AI system, disclosure of emotion recognition and biometric categorisation, and marking of synthetic content. Article 49 registration in the EU database took effect. National market surveillance authorities gained their enforcement powers, and the penalty regime became applicable to these obligations.