The AI Act, the first broad AI law, coming into force in stages, and the changes made to it.
Under the omnibus deal, systems placed on the market before 2026-08-02 have until this date to mark AI-generated content; the new ban on AI that generates non-consensual intimate images or CSAM has a transition to this date. The omnibus entered into force on 2026-07-27.
Sandbox set-up deadline moves to this date under the omnibus (Gibson Dunn summary of the agreed text).
High-risk AI in areas such as biometrics, education, employment and migration must meet the Act's requirements.
European Commission, AI Act page (Shaping Europe's digital future)
AI embedded in products covered by EU product-safety law must comply.
European Commission, AI Act page (Shaping Europe's digital future)
Laws, orders, regulator rules, court rulings and deadlines. Filter by kind.
Transparency rules apply, and the AI Office can request documentation, evaluate GPAI models, order fixes and impose fines.
“The AI Act entered into force on 1 August 2024 and became applicable on 2 August 2026”
European Commission, AI Act page (Shaping Europe's digital future), Oct 19, 2026
Delays stand-alone high-risk rules to 2027-12-02 and product-embedded high-risk rules to 2028-08-02, and adds a ban on AI generating non-consensual sexual content and CSAM from December 2026.
European Commission, AI Act page (Shaping Europe's digital future), Oct 19, 2026
Parliament and Council negotiators agreed to push stand-alone (Annex III) high-risk duties to 2027-12-02 and product-embedded (Annex I) duties to 2028-08-02, add a ban on 'nudifier' and CSAM-generating AI, and give the AI Office more enforcement powers. Member states confirmed it on 2026-05-13. (The Commission dates the political agreement 2026-05-07.)
Gibson Dunn (law firm alert), May 27, 2026
Part of the digital omnibus simplification package; proposed delaying high-risk obligations.
European Commission, AI Act page (Shaping Europe's digital future), Oct 19, 2026
Providers of general-purpose AI models must meet the Act's GPAI obligations.
European Commission, AI Act page (Shaping Europe's digital future), Oct 19, 2026
Voluntary code for GPAI providers. The Commission lists 22 signatories (AI Studio Delta, Aleph Alpha, Almawave, Amazon, Anthropic, Black Forest Labs, Bria AI, Cohere, Domyn, Dweve, Fastweb, Google, IBM, LINAGORA, Microsoft, Mistral AI, Open Hippo, OpenAI, Pleias, Reflection AI, ServiceNow, WRITER); xAI signed only the Safety and Security chapter.
“The code was published on July 10, 2025.”
European Commission, The General-Purpose AI Code of Practice, Oct 7, 2026
Three instruments to help general-purpose AI providers comply, one of them a voluntary code of practice.
European Commission, AI Act page (Shaping Europe's digital future), Oct 19, 2026
Prohibited practices and the AI-literacy obligation start to apply.
European Commission, AI Act page (Shaping Europe's digital future), Oct 19, 2026
The EU's AI law takes effect, with obligations phased in over the following years.
European Commission, AI Act page (Shaping Europe's digital future), Oct 19, 2026
Read at each source on Oct 19, 2026. All places
