Connecticut Passes SB5 — America's Most Comprehensive State AI Law
What Passed
Connecticut's legislature passed the AI Responsibility and Transparency Act (SB5) — House 131-17, Senate 32-4. Governor Ned Lamont is expected to sign. Most provisions take effect October 1, 2026.
Key Requirements
- Provenance metadata: Generative AI providers with 1M+ monthly users must embed machine-readable provenance data in all AI-generated audio, images, and video
- Whistleblower protection: Frontier model developers training on more than 10²⁶ computing operations must protect employees who report catastrophic risk concerns
- Employment AI: Transparency requirements for employers using AI in hiring and promotion
- AI companions: Disclosure obligations for AI emotional companion services
Significance
With no comprehensive federal AI law in force, SB5 is among the broadest state-level AI statutes in the US. The 1M-user threshold for provenance requirements is lower than prior proposals, capturing a wide range of consumer-facing generative AI products. The 10²⁶ FLOP whistleblower threshold mirrors the compute levels used in frontier foundation model training in 2025–2026.
Source: CT Mirror
Related Articles
The Trump administration is considering an executive order requiring government review of powerful AI models before release, with NSA and intelligence agencies as oversight bodies. The move marks a sharp reversal after revoking Biden's AI safety order in January 2025.
A 269-page House discussion draft released on June 4 would create a federal AI framework while blocking state laws aimed at AI model development for three years. The fight is already about whether national consistency is worth slowing state-level AI rules.
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.