The European Parliament votes to ban nudify apps and delay EU AI Act deadlines, including pushing compliance for high-risk AI systems back to December 2027
The proposals would push back looming deadlines for watermarking AI-generated content and high-risk AI systems.
Context & Ripple Effects
The vote reopens implementation terms for a regime that entered into force with staged compliance deadlines, after the Parliament's earlier AI Act push sought limits on facial recognition and disclosure obligations. It matters because the Parliament is pairing a targeted prohibition with a slower timetable for broader obligations.
The direction was later reflected in legislators' deal to move high-risk AI restrictions to December 2027, indicating that the timetable change was becoming a substantive revision to the law's rollout rather than a one-off procedural delay.
First-order effects
- Providers and deployers of high-risk AI gain more time before the proposed December 2027 compliance point, while the planned watermarking deadline also moves back.
- Nudify-app operators face a proposed EU-level ban, separating this use case from the broader extension of compliance timelines.
Second-order effects
- AI vendors can defer some near-term documentation, testing, and content-provenance work, while customers relying on high-risk systems receive a longer period of regulatory uncertainty.
- The split treatment pressures policymakers and companies to distinguish clearly between applications judged inherently harmful and systems subject to risk-based compliance rules.
Third-order effects
- If enacted, the change would make the AI Act's rollout more adaptive to implementation capacity: fast intervention for narrowly defined harms, slower enforcement for technically and operationally complex obligations.
- The eventual effectiveness of the regime will depend more heavily on enforcement definitions and timelines, as delayed watermarking and high-risk requirements reduce the period in which uniform obligations shape the market.
The trend: EU AI governance is shifting toward selective, harm-specific bans alongside a more phased timetable for broad risk-based compliance.