Connecticut's AI law is now partly in force. The Artificial Intelligence Responsibility and Transparency Act, Public Act 26-15, began taking effect on 1 October. AI subscription providers must disclose key terms in writing before signing or renewing a consumer. Generative systems with more than one million monthly users must include provenance data in audio, image and video where commercially and technically reasonable. Frontier developers may not adopt policies permitting retaliation against employees who report catastrophic risk. Employers filing WARN notices must tell the Labor Department whether layoffs relate to AI, and the use of automated employment technology cannot serve as a defence to a discrimination complaint.
The next instalment arrives on 1 January 2027, when AI companions need self-harm protocols and not-human notices and large frontier developers need anonymous internal reporting. From 1 October 2027 employers deploying automated decision tools owe plain-language and written pre-decision notice. The Attorney General holds enforcement for most provisions.
Sources: cga.ct.gov (Public Act 26-15); housedems.ct.gov (2 Oct 2026); natlawreview.com (30 Sep 2026); pewresearch.org (17 Jun 2026); cred-iq.com (1 Oct 2026); billboard.com (30 Sep 2026); news.pollstar.com (31 Aug 2026).