Connecticut AI Law Sets Oct 2026 Start, Holds Employers Accountable for Bias

TL;DR: Connecticut’s AI workplace law becomes enforceable on Oct 1, 2026, requiring employers to disclose AI use and prove tools are free of discriminatory impact.

Context Connecticut joins a growing list of states regulating artificial intelligence in hiring and other employment decisions. The legislation, known as the Artificial Intelligence Responsibility and Transparency Act (SB 5), is slated for the governor’s signature. It arrives amid a patchwork of state rules that demand transparency, risk assessments, and anti‑bias safeguards for automated hiring systems.

Key Facts - The law defines an “automated employment‑related decision process” (AERDP) as any computational output—score, rank, recommendation, or classification—that materially influences hiring, promotion, discipline, or termination. This broad definition captures resume‑screening software, AI chatbots, video‑interview analytics, automated background‑screening flags, performance‑rating engines, and workforce‑management tools. - Employers must provide plain‑language notices before using an AERDP. Disclosures must explain the tool’s purpose, the data it processes, the type of output it generates, and whether human review is possible. If an adverse decision follows, a separate statement must detail how the AI output contributed, what data were used, and the data source. - The law amends the Connecticut Fair Employment Practices Act to treat discriminatory outcomes from AI as the employer’s liability, not the vendor’s. Courts will consider the presence, quality, and recency of any anti‑bias testing when evaluating compliance. - Implementation begins on Oct 1, 2026, with staggered rollout for different provisions, giving companies time to audit existing tools and adjust processes.

What It Means Employers in Connecticut can no longer claim that an AI system shields them from discrimination claims. They must conduct statistical analyses to detect disparate impact on protected classes such as race, gender, or age. If disparities appear, companies must demonstrate that the tool is job‑related and consistent with business necessity, or modify or discard it.

Human resources departments will need to inventory every automated system that influences employment decisions, assess its data inputs, and document testing results. Vendors may be required to supply validation studies showing the tool’s fairness and accuracy. Failure to meet these standards could trigger lawsuits under state anti‑discrimination law.

The law also raises the bar for transparency. Candidates and employees will receive clear notices about AI involvement, potentially increasing scrutiny of hiring practices and prompting broader industry shifts toward explainable AI.

Looking ahead, watch for the first compliance audits in late 2026 and any legal challenges that could shape how other states adopt similar AI accountability frameworks.