Connecticut's AI Responsibility and Transparency Act takes effect October 1

A law-firm briefing on its site ebglaw.com reviews state and federal rules on workplace artificial intelligence, with Connecticut's SB 5 among the laws it covers. The firm says Connecticut's governor, Ned Lamont, signed the bill on 27 May 2026, listed as Public Act 26-15, after the General Assembly passed it on 1 May. According to the briefing, the law reaches AI-assisted employment decisions on hiring, promotion, discipline, training and material changes to employment terms, including termination, but leaves out minor matters such as changes in tasks, hours or assignments.

The briefing says some provisions start on October 1, 2026. From that date an employer carrying out a layoff that counts as a mass layoff under the federal WARN Act must give the state Department of Labor a written notice saying whether AI or other technology contributed. The firm stresses this applies to any such layoff or plant closing in the state, not only those where AI chose who would go.

Further duties follow. From October 1, 2027, employers using automated decision technology that deals directly with applicants or employees must tell them so in plain language, unless a reasonable person would find it obvious. The firm lists items a notice should contain, such as the purpose, the product's trade name, the personal data analysed and contact details. The law also bars employers from citing the technology as a defence to a discrimination claim, though courts may weigh evidence of bias testing. A grace period to 31 December 2027 lets the Attorney General give notice and a chance to fix problems before enforcing.

The briefing places Connecticut alongside Colorado, California, Illinois and Texas, and describes federal efforts to preempt state rules. It concludes that the patchwork of requirements is likely to keep growing and changing.