The compliance calendar changed on March 26.
The European Parliament adopted its position on a simplification proposal for the EU AI Act, shifting three major implementation deadlines and adding a new prohibition on AI nudifier systems. These aren’t final dates. The Parliament’s position now goes to the Council of the European Union for approval. But compliance teams can plan against them.
The three proposed new deadlines:
| Compliance Area | New Proposed Deadline | Who’s Affected |
|---|---|---|
| AI-generated content watermarking | November 2, 2026 | Providers of AI systems generating audio, image, video, or text |
| High-risk AI systems (listed) | December 2, 2027 | Providers of high-risk AI systems in Annex III of the regulation |
| Sectoral AI systems | August 2, 2028 | Providers of AI systems under EU sectoral safety and market surveillance law |
The watermarking deadline is the most pressing. At seven months out, it’s the nearest obligation on this revised calendar and applies broadly, any provider generating AI content for EU audiences should be assessing its watermarking readiness now, not after Council approval.
The nudifier ban is a new prohibition, not a delay. The Parliament’s position proposes banning AI systems that create or manipulate sexually explicit images resembling identifiable real persons without their consent. This is an expansion of the regulation’s scope, added in the same vote that extended the implementation timeline. Delay and expansion in the same action is a signal about Parliament’s intent: the framework is getting broader, not narrower.
What still has to happen. Parliament’s position is adopted, but it isn’t law. The Council of the EU must approve the proposal before any of these dates become binding. Parliament stated the proposed delays are intended to ensure that guidance and technical standards are in place before companies must comply, an acknowledgment that the original timeline outpaced the infrastructure needed to meet it.
Why the watermarking deadline matters most right now. November 2026 is seven months away. High-risk system compliance (December 2027) and sectoral compliance (August 2028) provide longer runways. But watermarking affects a wide category of providers, and technical implementation, building systems to label AI-generated audio, image, video, and text, takes time to scope, build, and test. Organizations generating AI content for EU audiences should treat November 2026 as the planning horizon, not the deadline.
What to watch. The Council’s response to Parliament’s position is the next milestone. Until the Council approves, these dates are proposed targets, not enforceable deadlines. Watch also for the guidance and standards Parliament referenced, those documents will define what “watermarking compliance” and “high-risk system compliance” actually require in practice. The official press release is the authoritative source for the current proposed dates.
For the full analytical picture of what the EU deadline shift means alongside the US deregulatory push, see the companion deep-dive this cycle: US Deregulates, EU Extends: What the Same-Week Moves From Washington and Brussels Mean for Global AI Compliance.