EU publishes AI Act Omnibus with new deadlines and a ban on non-consensual AI intimate images
LuizaJarovsky · x · 2026-07-24
The EU has published the Digital Omnibus amending the AI Act in the Official Journal, with several substantive changes.
Key changes
- New timelines: high-risk systems under Annex III move to 2 Dec 2027; Annex I systems to 2 Aug 2028; AI regulatory sandboxes must be in place by 2 Aug 2027; transparency duties for AI-generated content already on the market get a grace period ending 2 Dec 2026.
- New banned practice: non-consensual sexual/intimate content and CSAM generated by AI are prohibited; nude-image generation or removing clothes from real people’s photos is banned starting Dec 2026.
- AI Office scope clarified: the EU AI Office will supervise certain GPAI-based systems when both the model and the downstream system are developed by the same provider, while some domains remain under national authorities.
- Sector rules and machinery exemption: the Omnibus narrows overlap with existing sectoral laws and exempts machinery-regulated products from direct AI Act applicability, while empowering the Commission to add safety requirements under machinery rules.
- Compliance burden: the Commission must provide guidance to reduce compliance burden for Annex I high-risk systems already covered by sectoral regulation.
More from Safety
- NYT roundtable asks five experts what the rules for AI companies should actually be — ShakeelHashim · 2026-10-26
- Google DeepMind pilots world's first double-blind evaluations for frontier models — scychan_brains · 2026-10-26
- LLM back-of-envelope estimate puts PHP vulnerability impact at $1bn–$4bn — S_OhEigeartaigh · 2026-10-26
- US states halt new data centers as 72% of Americans oppose local builds — YvesMulkers · 2026-10-26
- LEMON: zero-shot LLM misbehavior monitor flags hallucinations from attention maps — yoavartzi · 2026-10-26
- a16z podcast: Diode pairs AI PCB design with Brooklyn factory to revive US hardware manufacturing — MattPerault · 2026-10-26