Connecticut passes AI regulation after years of stalled efforts
Connecticut's House of Representatives approved comprehensive artificial intelligence legislation Friday, sending Senate Bill 5 to Gov. Ned Lamont's desk with bipartisan support. The House voted 131-17 in favor. The Senate had passed it 32-4 the previous week.
Lamont's office said he plans to sign the bill into law, clearing the final hurdle after years of failed attempts. The legislation would make Connecticut one of a growing number of states regulating AI as the technology expands faster than existing rules can address.
What the bill covers
Rebranded as the Connecticut Artificial Intelligence Responsibility and Transparency Act, the bill addresses employment decisions made by AI systems, AI use within state agencies, and regulations around AI chatbots and social media interactions involving minors.
The law also funds workforce development initiatives, including the Connecticut AI Academy and AI literacy programs for teachers and small business owners.
A regulatory sandbox provision allows companies to test new AI technologies under state oversight-a provision that helped secure governor support.
How this differs from previous attempts
Last year, similar legislation cleared the Senate but died after Lamont threatened to veto it. He had worried that strict rules would discourage innovation and hurt Connecticut's business climate.
This year's version incorporated provisions from multiple bills, including some the governor had requested. That compromise-brokered between Lamont and the bill's lead author, Sen. James Maroney-proved decisive.
"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 Rep. Roland Lemar, a Democratic co-chair of the General Law Committee, during Friday's debate. "This is about protecting people without stopping that innovation."
Business concerns addressed
Several Republican House members who had worried about regulatory overreach said their concerns had been addressed. Rep. David Rutigliano, R-Trumbull, said the bill sets necessary boundaries without stifling economic development.
Not all lawmakers agreed on the approach. Rep. Bill Buckbee, R-New Milford, called parts of the law "theater," arguing the state cannot effectively regulate technology that changes too quickly. He noted he had used AI to generate questions he asked during the debate itself.
Rep. Tina Courpas, R-Greenwich, raised concerns about Connecticut's ability to compete with other states if regulations become too restrictive.
Federal policy uncertainty
Connecticut's move comes as the Trump administration has pushed states to avoid AI regulations, arguing that federal policy should create a single national standard instead.
Connecticut Attorney General William Tong supported the bill's passage. "Artificial intelligence is exploding everywhere, with potential for good intertwined with severe risks," he said. "Neither state nor federal law has kept pace with these developments."
A companion bill on consumer data privacy, Senate Bill 4, is expected to come up for a vote soon.
For IT and development professionals, understanding state-level AI regulations has become essential. AI for IT & Development covers how these rules affect implementation and deployment. Connecticut's focus on Generative AI and LLM systems means development teams will need to understand compliance requirements around chatbots and automated decision-making systems.
Your membership also unlocks:
