EU AI Act · Timeline
EU AI Act Timeline: Key Dates and Deadlines
Last updated: 26 August 2026
The EU AI Act is being phased in over several years. Here's the full timeline: what already applies, what's coming, and what's been deferred.
See also our full guide to the EU AI Act for the context behind each date.
1 August 2024: AI Act enters into force
Status: already applies.The EU AI Act formally becomes law, starting the phased timeline for when its various rules take effect. Entry into force itself doesn't yet create concrete obligations for most businesses: it marks the start of the rollout.
2 February 2025: AI literacy and prohibited practices
Status: already applies. From this date, businesses that provide or use AI systems are expected to ensure a reasonable level of AI literacy among their staff. At the same time, the ban on a short list of particularly intrusive AI practices takes effect, including social scoring and certain manipulative techniques.
2 August 2025: General-purpose AI and governance
Status: already applies.Rules for providers of general-purpose AI models become applicable, alongside the Act's governance structure, including the national and European authorities responsible for overseeing the rules.
2 August 2026: Transparency obligations
Status: applies now. Certain AI systems must make it clear that people are dealing with AI, and AI-generated or manipulated content (including deepfakes) must in certain cases be labelled. These rules (Article 50) were notdeferred by the later “Digital Omnibus” simplification. The European Commission adopted its final guidelines on Article 50 on 20 July 2026, shortly before the obligations took effect. Read the full breakdown on our transparency obligations page.
2 December 2027: Most high-risk AI systems (deferred date)
Status: expected from this date.Requirements for most high-risk AI systems (standalone systems under the Act's Annex III, e.g. in employment and credit scoring) were originally due to apply from 2 August 2026. The date was deferred to 2 December 2027 as part of the “Digital Omnibus” simplification package, which the Council gave final approval on 29 June 2026, following the European Parliament's endorsement on 16 June 2026.
2 August 2028: High-risk AI in regulated products (deferred date)
Status: expected from this date.High-risk AI systems embedded in products already regulated under other EU law (such as medical devices and machinery, under the Act's Annex I) were originally due to apply from 2 August 2027. The date was deferred by 12 months as part of the same simplification package.
What could still change?
The timeline above reflects the adopted “Digital Omnibus” simplification and published European Commission guidance, but implementation details (further guidance, standards, or minor adjustments) may still follow. We update this page as official changes are published.
Sources
This page is general information, not legal advice. Prepared and reviewed by Complaia.