EU Parliament and Council Strike Deal to Simplify AI Act, Delay High-Risk Rules to Late 2027

Overview

The European Parliament and EU Council reached a provisional agreement on May 7 to amend the AI Act, substantially extending compliance deadlines for high-risk AI systems while adding new prohibitions targeting AI-generated harmful content.

Revised Deadlines

• Stand-alone high-risk AI systems: December 2, 2027
• High-risk AI embedded in regulated products: August 2, 2028
• National AI regulatory sandboxes: August 2, 2027
• Synthetic content transparency solutions: December 2, 2026 (grace period cut from 6 to 3 months)

Key Changes

The deal extends simplified documentation requirements — previously limited to SMEs — to small mid-cap companies (SMCs) with up to 500 employees. It centralizes enforcement in the AI Office, reducing fragmentation across member states, and creates a new EU-level regulatory sandbox.

New Prohibitions

Co-legislators added a provision prohibiting AI practices that generate non-consensual sexual and intimate content, including child sexual abuse material (CSAM) — going beyond the original Commission proposal.

Next Steps

The agreement is provisional and requires formal adoption by both institutions. For businesses preparing for an August 2026 deadline, the revised schedule provides at least 16 months of additional runway.

Source: EU Council Press Release