Daily Digest: May 7, 2026

Top Stories – May 7, 2026

EU Council and Parliament Agree to Delay High-Risk AI Act Rules Until December 2027 Tier 2

The EU Council and European Parliament reached a provisional agreement today pushing the enforcement deadline for high-risk AI systems – including biometrics, law enforcement, and critical infrastructure – from August 2026 to December 2027, citing business compliance costs and competitiveness concerns. The deal also expands regulatory sandboxes and removes overlapping obligations for machinery already governed by sector-specific legislation; it still requires formal approval by member states and Parliament.

Connecticut Passes Comprehensive AI Regulation, Governor to Sign Tier 2

Connecticut’s legislature passed a wide-ranging AI bill with strong bipartisan margins – 131-17 in the House and 32-4 in the Senate – and the governor has indicated he will sign it. The legislation covers frontier model governance, chatbot disclosure, employment protections, and content provenance requirements, making Connecticut one of the most active states on comprehensive AI law.

Colorado’s AI Law Rewrite Would Drop Algorithm Explainability Requirements Tier 2

Colorado’s SB 26-189 advances a near-total replacement of SB 24-205 – the nation’s first consumer AI protection law – and would eliminate the requirement that companies explain how their AI systems reach consequential decisions. The narrowed framework pushes the effective date to January 2027 and significantly scales back the algorithmic accountability obligations that made the original law a national benchmark.