Connecticut Passes the Broadest AI Law in the US: What It Requires Companies to Disclose About Chatbots and Subscriptions

While the US federal government pushes for AI deregulation nationwide, Connecticut passed one of the broadest state-level AI laws to date. SB 5 doesn't just regulate consumer chatbots — it also touches frontier models, automated employment decision tools, and, in a detail almost no one saw coming, AI software subscriptions themselves.

What the law covers and when it takes effect

Connecticut's General Assembly passed SB 5 on May 11, 2026, and Governor Ned Lamont signed it on May 27. According to analysis from firms like WilmerHale, Morrison Foerster, and DLA Piper, the law imposes obligations on consumer-facing AI chatbots, frontier model developers, developers and deployers of automated employment decision tools, and social media platforms. Its compliance timeline is staggered, with effective dates running from October 2026 through January 2028.

The surprising clause: transparency on AI subscriptions

The contrast with the federal stance

This law arrives amid a growing divergence: while the United States pushes for a looser approach to AI regulation at the federal level — including a December 2025 executive order aimed at blocking state laws deemed incompatible with a minimal national framework — states like Connecticut are moving in the opposite direction. Nearly 100 chatbot-specific bills have been introduced across 34 states in 2026, suggesting Connecticut's SB 5 will be a reference model, not an isolated case.

What it means for companies selling or using subscription-based AI