Connecticut mandates AI system inventory and ongoing assessments by December 2023

Connecticut's Senate Bill 2, introduced on April 9, 2025, aims to enhance oversight and accountability of artificial intelligence (AI) systems utilized by state agencies. The bill mandates an annual inventory of all AI systems in use, requiring detailed documentation on each system's capabilities, vendor information, and whether it has undergone an impact assessment prior to implementation. This inventory will be made publicly accessible through the state's open data portal, promoting transparency.

A key provision of the bill is the establishment of ongoing assessments to ensure that AI systems do not lead to unlawful discrimination or disparate impacts. These assessments will be conducted by the Commissioner of Administrative Services, following guidelines set by the Office of Policy and Management. The bill also emphasizes the importance of training for state employees on the responsible use of generative AI, developed in consultation with various stakeholders, including collective bargaining units and industry experts.

The introduction of Senate Bill 2 has sparked notable discussions among lawmakers and advocacy groups. Proponents argue that the bill is a necessary step toward responsible AI governance, addressing concerns about bias and discrimination in automated decision-making processes. Critics, however, express concerns about the potential bureaucratic burden on state agencies and the feasibility of implementing comprehensive training programs.

The implications of this legislation are significant, as it positions Connecticut as a leader in AI regulation, potentially influencing similar initiatives in other states. Experts suggest that the bill could set a precedent for how public sector entities manage AI technologies, balancing innovation with ethical considerations.

As the bill moves through the legislative process, its outcomes will be closely monitored, particularly regarding its impact on state operations and the broader conversation about AI ethics in governance. The next steps will involve further debates and potential amendments as lawmakers seek to refine the bill's provisions before a final vote.