Connecticut Governor signs SB 5, mandating AI provenance and disclosures

The law imposes immediate technical requirements for content watermarking and provenance that will impact video production and distribution workflows. For the streaming ecosystem, this accelerates the shift toward mandatory transparency in synthetic media, moving beyond voluntary industry standards like C2PA. The bipartisan consensus in Connecticut signals that state-level regulation is pivoting từ safety theory to granular operational mandates for data transparency and vendor management. Industry leaders should watch the October 1, 2026 effective date for initial governance updates, which will serve as a bellwether for similar pending legislation in other key markets.

Connecticut’s enactment of SB 5 follows years of executive-legislative tension. Governor Lamont previously threatened to veto similar measures in 2024 and 2025, citing concerns that premature state-level regulation would stifle the local technology sector, per the Hartford Business Journal (July 2025). The 2026 compromise, re-titled the 'Connecticut Artificial Intelligence Responsibility and Transparency Act,' gained bipartisan support by focusing on specific consumer and workforce protections rather than the broader developer liability found in other states’ proposals. This legislative shift occurred as several other states refined their own approaches; for instance, Colorado passed the landmark SB 205 in May 2024, set to take effect in February 2026, which focuses heavily on preventing algorithmic discrimination in 'high-risk' systems.

While state-level activity accelerates, federal lawmakers are attempting to reassert authority through preemption. According to FedScoop (June 2026), a bipartisan House group led by Representatives Lori Trahan and Jay Obernolte recently unveiled the 'Great American AI Act.' This 269-page discussion draft proposes a three-year sunset on state laws that specifically regulate the development of AI models—a move that would target transparency requirements like California’s AB 2013 and parts of Connecticut’s synthetic media rules. Meanwhile, California Governor Gavin Newsom has taken a bifurcated approach, vetoing a high-profile safety bill (SB 1047) in late 2024 while approving several transparency-focused laws regarding training data and deepfake detection, according to reports from Morgan Lewis (October 2024).

The Connecticut law distinguishes itself by specifically targeting 'AI companions' used for romantic or social interactions, following a model seen in Idaho and Iowa, per DLA Piper (May 2026). Unlike New York City’s Local Law 144, which requires independent bias audits for employment AI, Connecticut’s SB 5 mandates transparency and notice but stops short of requiring third-party testing. These variations indicate that video and technology firms must now navigate a highly fragmented compliance landscape where disclosure requirements vary significantly by jurisdiction and specific AI use case.