AI Frontier Institute

An Act Addressing Innovations In And The Responsible Use Of Artificial Intelligence.

View primary source →

Connecticut Substitute Bill No. 86 has not been enacted. It was reported out favorably by committee and tabled on the Senate calendar (Calendar Number 225). If enacted, it would do three things. (1) Section 1 amends section 4-67p of the general statutes. By January 1, 2028, the state Chief Data Officer, working with agency data officers, would review agencies' high-value data inventories. The goal is to identify and publish data useful for AI, machine learning and other statistical analysis, to support economic opportunity and development. The officers would identify suitable data. They would set data quality and governance policies that avoid unlawful discrimination or disparate impact. They would decide what aggregation or redaction is needed to protect privacy. They would also set procedures for publishing the data on the state's online open-data repository. This takes effect July 1, 2027. (2) Section 2 directs the Commissioner of Economic and Community Development, in consultation with the Banking, Administrative Services, Public Health and Insurance commissioners, to develop a plan for an AI regulatory sandbox. The sandbox would let applicants temporarily test innovative products or services on a limited basis under reduced licensure and regulatory requirements. Recommendations for any needed legislation are due to the Governor and the relevant committees by January 1, 2028. (3) Section 3, a new section effective January 1, 2027, regulates operators of companion chatbot platforms. Operators must give a clear notice that a chatbot is artificial if it could mislead a reasonable person into thinking it is human. Operators must also keep and publish a protocol to prevent chatbot content about suicide or self-harm and to refer users who express suicidal ideation or self-harm intent to crisis services. They must disclose that companion chatbots may not suit some minors. For users known to be minors, operators must disclose the AI nature of the chatbot and prompt a break at least once after each three hours of continuous use. They must also take reasonable measures to prevent sexually explicit visual depictions or direct statements urging sexually explicit conduct. Starting February 1, 2028, operators must report annually to the Attorney General on their protocols and the number of crisis referrals, without including personal data. The Attorney General must post these reports online. Violations are unfair trade practices enforced solely by the Attorney General, and there is no private right of action.

← Back to the tracker