Connecticut Sends a Sweeping AI Bill to Lamont's Desk

Connecticut's General Assembly has passed Senate Bill 5, the AI Responsibility and Transparency Act, in a deal that sends the most comprehensive state-level AI bill in the country to Governor Ned Lamont's desk. The House passed it 131 to 17 on May 1. The Senate cleared it 32 to 4 in late April. A spokesperson for the governor told the Connecticut Mirror that Lamont looks forward to signing it.

The bill is broader in scope than the AI laws already on the books in California and Colorado. It requires the largest generative AI providers, those with more than a million monthly users in the state, to embed machine-readable provenance data in any audio, image, or video their systems generate. It imposes safety obligations on operators of consumer chatbots, with stronger requirements when the chatbot interacts with minors. It creates whistleblower protections for employees at frontier-model developers, defined as those training models on more than 1026 computing operations. And it requires employers issuing layoff notices to disclose to the state whether the cuts are connected to AI deployment.

Several provisions take effect on October 1, 2026, if signed. Others, including the synthetic content rules, phase in through 2027.

The road to passage was long. A similar bill cleared the Senate in 2025 but died after a veto threat from Lamont. This year's version was negotiated to include a regulatory sandbox and youth social media protections that secured his support. The resulting law is, in DLA Piper's words, one of the most comprehensive AI laws in the country.

Connecticut's move is likely to influence other states. With federal action stalled after the White House quietly shelved its AI safety executive order in late May, and after the EU spent the spring simplifying and delaying parts of its own AI Act, the legal centre of gravity around AI is shifting back to the statehouses.