The bill's effective dates are staggered across late 2026 and beyond, with specific implementation timelines not yet finalized in available sources. The scope of "automated employment-related decision systems" and the precise contours of the disclosure and testing requirements will become clearer as regulatory guidance emerges.
Connecticut joins California, Colorado, New York City, and other jurisdictions tightening AI governance in employment contexts. Critically, the statute explicitly prohibits automated systems from serving as a defense to discrimination claims—meaning employers cannot shield themselves from liability by pointing to algorithmic decision-making. Evidence of bias testing and anti-discrimination protocols may reduce exposure but will not eliminate it. Employers deploying AI for HR decisions should immediately audit their tools, review vendor contracts for compliance gaps, and establish human-review and bias-testing protocols before the law takes effect.