--- headline: "Connecticut Passes One of America's Most Comprehensive AI Laws Covering Employment, Youth, and Chatbots" slug: connecticut-ai-responsibility-act category: policy story_number: "11" date: 2026-05-07 ---
# Connecticut Passes One of America's Most Comprehensive AI Laws Covering Employment, Youth, and Chatbots
Connecticut has become the latest state to draw a regulatory line around artificial intelligence, passing a sweeping 71-page law that reaches across employment discrimination, youth safety, chatbot transparency, and whistleblower protections for workers at the largest AI companies. The Connecticut Artificial Intelligence Responsibility and Transparency Act cleared the state House of Representatives by a vote of 131 to 17 on May 1 after passing the Senate 32 to 4, and Governor Ned Lamont has confirmed he will sign it.
The bipartisan margins are notable in a policy area where state legislatures have struggled to build consensus. Connecticut spent years failing to pass AI regulation, with a previous version clearing the Senate in 2025 only to collapse after Lamont threatened a veto over concerns that heavy-handed rules would drive tech companies out of the state. This time, a deal between the governor and the bill's lead author, Senator James Maroney of Milford, produced a measure that both sides could accept.
"There is no longer doubt that the nature of work, the nature of life, is going to change rapidly with the continued evolution of AI," said Representative Roland Lemar, a New Haven Democrat who co-chairs the General Law Committee. "This is about protecting people without stopping that innovation."
What the Law Covers
Senate Bill 5 is not a single-issue regulation. It operates across several distinct domains, each with its own compliance timeline beginning October 1, 2026, with additional provisions taking effect January 1, 2027, and October 1, 2027.
Employment. Developers of AI tools used as a "substantial factor" in hiring, promotion, discipline, or termination must provide deployers with compliance-related information. Employers must notify job applicants and employees when automated decision technology is being used, disclose the purpose and data categories involved, and explain data sources. Using AI tools is explicitly not a defense against discrimination claims under state law. Connecticut small businesses are shielded from liability if a third-party vendor introduces AI into the hiring process without their knowledge.
Youth and chatbot safety. Operators of AI companions, defined as systems that can sustain a relationship across multiple interactions, must prevent the chatbot from claiming to be human, including when a user directly asks. The law requires operators to implement measures that meet or exceed industry standards to prevent harmful interactions with minors, including romantic engagement. AI companions must use evidence-based methods to detect user expressions showing risk of suicide, self-harm, or imminent violence, and must refer users to appropriate mental health resources including the 988 Suicide and Crisis Lifeline.
Frontier AI whistleblower protections. In a provision with national implications, the law creates protections for employees of developers who train foundation models using more than 10 to the 26th power computing operations. These workers cannot be fired, disciplined, or penalized for reporting activity that poses a specific and substantial danger to public health or safety through catastrophic risks, including expert-level assistance in creating chemical, biological, radiological, or nuclear weapons, or unmonitored cyberattacks causing serious harm.
State agency oversight and workforce development. The law requires state agencies to inventory their AI systems and conduct impact assessments. It creates a Connecticut AI Academy to expand AI literacy among unemployed workers, small businesses, teachers, and state workforce programs. AI will be added to K-12 computer science curricula. Employers conducting mass layoffs must disclose when those layoffs are related to AI or other technology.
Political Context
Attorney General William Tong praised the bill's passage, framing the moment in terms of regulatory urgency. "Artificial intelligence is exploding everywhere, with potential for good intertwined with severe risks," Tong said in a statement. "Neither state nor federal law has kept pace with these developments, and today's vote is an important first step towards protecting Connecticut families."
Not everyone was convinced. Some Republicans warned the bill could restrain innovation. But even skeptics found reasons to vote yes. The legislation arrives as states increasingly move to fill a federal vacuum. The Trump administration has urged states not to pass their own AI rules, arguing for a single national standard, but Washington has produced no comprehensive federal AI law. Connecticut joins Colorado, which passed its own AI anti-discrimination law in 2024, in the small but growing group of states building their own regulatory frameworks.
The Workforce Dimension
The jobs question looms over the entire debate. Lamont's office cited data showing that since August 2024, nearly 11,000 job postings in Connecticut have required AI skills, a 40 percent increase from the prior year. One in 52 jobs in the state now lists AI skill requirements, up from one in 70 a year earlier, with the demand concentrated in roles requiring an associate or bachelor's degree.
The law attempts to address both sides of this transformation, promoting AI workforce training while requiring transparency when AI eliminates jobs. The disclosure requirement for AI-related mass layoffs is a provision that few other states have attempted.
What to Watch Next
Governor Lamont is expected to sign the bill in the coming days. Once enacted, the staggered effective dates mean employers and AI developers will have months to prepare for compliance, but the clock starts ticking on October 1, 2026. A companion bill, Senate Bill 4, addressing consumer data privacy protections related to AI, is expected to come up for a separate vote. Legal analysts at Freshfields have noted that the law's breadth, covering employment, youth safety, chatbot behavior, and frontier AI development in a single statute, makes it one of the most comprehensive state-level AI laws in the country. Whether it becomes a model for other states or a cautionary tale about regulatory complexity will depend on how smoothly its staggered provisions translate from statute to practice.
“There is no longer doubt that the nature of work, the nature of life, is going to change rapidly with the continued evolution of AI. This is about protecting people without stopping that innovation.”— Roland Lemar, State Representative, Connecticut