An act intended to enhance online safety cleared the House Friday by a bipartisan 131-17 vote. The legislation, Senate Bill 5, passed the state Senate by a 32-4 vote last month and now heads to Gov. Ned Lamont.
The bill establishes protections against Artificial Intelligence harms, expands workforce development programs, and aims to position Connecticut as a leader in responsible Artificial Intelligence policy. Supporters said the measure balances resident protections with innovation, with provisions focused on shielding vulnerable people from dangerous chatbots, ensuring workers are treated fairly, and improving statewide Artificial Intelligence literacy.
Republican support helped drive the bill forward. State Rep. Joe Hoxha of Bristol said he agreed with most or all of the legislation after reviewing it. Hoxha, who had asked for a cost estimate on the bill, said the estimated ? million to ? million was worth it for the protections it included. He argued that individual judgment alone is not enough when technology makes real and fake difficult to distinguish.
State Rep. David Rutigliano of Trumbull pointed to a requirement that Artificial Intelligence companion operators include a protocol for detection of certain user expressions such as self harm. He said the state was not trying to suppress innovation or economic development, but to establish parameters for Artificial Intelligence programs that interact with children.
Senate Majority Leader Bob Duff said the legislation shows states do not need to wait for federal action to move on behalf of residents. Supporters framed the bill as a combined effort to protect children, safeguard workers, and help build an Artificial Intelligence-ready workforce in Connecticut.
