The EU AI Act did not become fully applicable on one day. By 25 October 2026, multiple phases had taken effect, while important high-risk system duties were scheduled for later dates after the 2026 simplification.
Read the phased timeline correctly
The AI Act entered into force on 1 August 2024. Its AI literacy provisions and certain prohibitions started to apply on 2 February 2025. Governance and general-purpose AI model obligations began applying on 2 August 2025. The main application milestone, including specified transparency provisions, arrived on 2 August 2026. These dates do not mean every system is regulated identically.
The 2026 amendment matters
EU institutions adopted AI Act simplification measures in 2026. European Commission materials checked on 25 October 2026 show delayed high-risk AI dates: 2 December 2027 for Annex III stand-alone systems and 2 August 2028 for systems embedded in the relevant regulated products under Annex I. Another transition date of 2 December 2026 applies to specified transparency measures and new prohibitions. Always verify the relevant provision and transitional conditions against consolidated law.
Determine your actual role
The Act distinguishes providers, deployers, importers, distributors and other actors. It also distinguishes prohibited practices, transparency duties, general-purpose model obligations and high-risk systems. Labeling a product “AI” does not establish that it meets the legal definition or occupies one specific category. Document intended purpose, system context, affected people and the entity’s role.
What users should do today
Keep an inventory of systems, assign compliance owners, train relevant staff, review disclosures for interactions and synthetic media, and assess whether high-risk categories or sectoral rules will become applicable. Continue meeting preexisting data-protection, consumer and safety duties. Avoid promising customers that a generic badge proves AI Act conformity.
Example: recruitment software
An organization using AI in employment should examine the actual function and Annex III category rather than assuming all chatbots are high-risk. The delayed date for certain high-risk obligations does not mean discrimination, privacy or labor rules are suspended. Obtain legal review in the country of deployment.
Sources and status warning
The original 2024 law’s Article 113 dates should not be quoted in isolation after the 2026 amendment. The Commission’s updated implementation timeline and regulatory framework page are the controlling editorial references for this dated explanation; reconfirm before publishing updates.
Dated first-party verification and limits
The Commission’s amended timeline distinguishes the 2 August 2026 transparency start from the limited 2 December 2026 transition for pre-existing systems under Article 50(2). New prohibitions covering specified non-consensual intimate and child sexual abuse material also have a 2 December 2026 start. For this 25 October 2026 snapshot, those December changes have not yet taken effect; Annex III and Annex I high-risk dates remain later milestones.
Sources & Further Reading
- EU AI Act implementation timeline after 2026 amendments — European Commission AI Act Service Desk.
- EU AI Act and simplification — European Commission.
- Regulation (EU) 2024/1689 — EUR-Lex.
Guidance varies by jurisdiction and application. This is general education, not legal, medical or other professional advice.