Consumer Protection Issue
By Barbara Heimlich
Editor
Emilia Otte, Connecticut Mirror contributed to this story
An Artificial Intelligence regularity bill had a bipartisan pass late Wednesday night in the Connecticut Senate.
Sen. James Maroney, D-Milford, who has become the General Assembly’s acknowledged expert on data privacy and AI, declined to speculate if the changes were sufficient to avoid a veto of Senate Bill 2.
Until Wednesday night, the dynamics this year were little changed from 2024, when the Senate Democratic majority passed a similar bill on a party-line vote, only to see it die from inaction in the House at the urging of the governor. Last year’s bill was viewed as a premature effort to set standards for the development and use of artificial intelligence and that it could hamper innovation and investment.
Sen. Paul Cicarella of North Haven, the ranking Senate Republican on General Law, said the revised measure maintains significant consumer protections while limiting the exposure of developers and users to litigation.
“We want businesses to come to Connecticut, know that they’re welcome, know we’re going to be good to them and not put burdens on them, whether that comes in the form of regulations or comes in the form of fees,” Cicarella said.
With the revisions drafted by Maroney and co-sponsored by four Republican senators, SB 2 passed on a 32-4 vote shortly before midnight, with all 25 Democrats and seven of the 11 Republicans in support.
It was unclear if the changes would reassure Lamont that state regulation would not discourage innovation involving artificial intelligence, a technology his chief economic adviser likens to a second industrial revolution.
“I just worry about every state going out and doing their own thing, a patchwork quilt of regulations, Connecticut being probably stricter and broader than most, what that means in terms of AI development here,” Lamont said earlier Wednesday.
The Connecticut bill passed Wednesday night requires greater transparency about uses of AI and makes a crime of disseminating “deep-fake porn,” synthetically produced intimate photos of an actual person.
The revisions strike a reference to “algorithmic discrimination.” That had been defined as “any use of an artificial intelligence system that results in any unlawful differential treatment or impact that disfavors” any protected class. It also exempts the use of AI in health care from the bill.
AI and algorithms are used widely to evaluate credit, job and housing applications, with evidence that some algorithms acting disfavorably on names deemed to be “Black sounding,” Maroney told reporters hours before the debate. “There are real harms that are happening,” Maroney said.
Sen. Gary Winfield, D-New Haven, complained about the deletion of algorithmic discrimination, which he described as an important provision to him and other members of the Black and Puerto Rican Caucus.
“I’m voting for this bill, but let’s not fool ourselves. This bill has a glaring hole in it, and it’s the same hole that just about every policy we do in this building has, as a whole,” Winfield said. “But broken people and broken lives are often the cost of business. I just hope we don’t break too many people’s lives.”
Sen. Tony Hwang, R-Fairfield, one of the four senators to oppose the bill, said even with the changes it was premature, rushed and overreaching.
“I am wholly supportive of AI and technology innovation, but it is a technology and innovation in which we need to watch and nurture, not try to watch and over regulate, because it will hurt our businesses,” Hwang said. “We don’t need to be the first. We need to be the best.”
Senate President Pro Tem Martin M. Looney, D-New Haven, and Senate Majority Leader Bob Duff, D-Norwalk, staged a press conference with Maroney prior to the debate to underscore that SB 2 was a priority of their caucus.
To read SB2 go to: