EU Agrees on AI Act Omnibus: High-Risk AI Deadlines Extended to 2027-2028, 'Nudification' Apps Banned
European Parliament and Council negotiators reached a provisional agreement on the Digital Omnibus regulation at 4:30 a.m. on May 7, 2026, closing six months of negotiations. The package amends the EU AI Act primarily by extending compliance deadlines for high-risk AI systems and adding new prohibitions on non-consensual sexual content generation.
Extended Deadlines for High-Risk AI
High-risk stand-alone AI systems now face a compliance deadline of December 2, 2027, and high-risk AI systems embedded in products must comply by August 2, 2028 — delays of roughly 16 and 24 months respectively from the original August 2, 2026 deadline. The rationale: harmonized standards, designated notified bodies, and conformity assessment procedures will not be ready in time.
New Prohibition: Non-Consensual Sexual AI Content
Co-legislators added a new prohibited AI practice covering non-consensual intimate imagery and child sexual abuse material (CSAM) generation. Companies must bring relevant systems into compliance by December 2, 2026 — sooner than most other obligations under the Omnibus.
Other Provisions
The deadline for AI-generated content transparency measures (watermarking) moves from 6 months to 3 months, targeting December 2, 2026. Simplified compliance is extended from SMEs to small mid-caps. The AI Act's risk-based architecture remains intact. Source: EU Council.
Related Articles
The EU AI Act’s transparency rules started applying on August 2, 2026, forcing many AI systems to disclose when users are interacting with AI or seeing AI-generated or edited content. Non-compliance can trigger fines of up to €15 million or 3% of global annual turnover.
The European Parliament and Council agreed on May 7 to simplify the AI Act, pushing high-risk compliance deadlines to December 2027 and August 2028 while adding a new ban on AI-generated non-consensual intimate imagery.
The EU has moved the heaviest high-risk AI Act obligations out of August 2026 and into two later dates: December 2, 2027 and August 2, 2028. The same package still tightens near-term bans and transparency deadlines around harmful generated content.