In Early May, by a 3–2 Margin, Hill County Officials Enacted a One Year Moratorium on New Data Center Construction in Unincorporated Areas

Why It Matters

The law forces employers to demystify AI decision‑making, reducing legal risk and aligning with a national push for algorithmic transparency in the workplace.

Key Takeaways

• Connecticut's SB5 requires AI notice for hiring, discipline, termination decisions.
• AI use disclosure required by Oct 1 2027 for hiring decisions.
• WARN‑type AI layoff notice required by Oct 1 2026.
• Exemptions cover word processors, spreadsheets, spam filters, and safety monitoring.
• Violations treated as unfair trade practices; AG enforces, no private right.

Pulse Analysis

Connecticut’s AI employment law joins a growing chorus of state‑level regulations aimed at curbing opaque algorithmic decision‑making. Following California, Illinois and Washington, the Constitution State’s SB 5 sets a clear benchmark: any tool that materially shapes hiring, discipline or termination outcomes must be disclosed to the affected individual. By anchoring the requirement to a specific date—Oct 1, 2027—the statute gives businesses a finite window to audit their tech stack, while the earlier WARN‑type notice for AI‑driven layoffs, effective Oct 1, 2026, signals that workforce reductions are under heightened scrutiny.

The bill’s definition of automated employment‑related decision technology (AEDT) is deliberately narrow, excluding everyday applications such as word processors, spreadsheets, spam filters and antivirus programs, as well as non‑material scheduling or safety‑monitoring functions. This carve‑out aims to avoid overburdening firms with trivial disclosures while still targeting sophisticated analytics, predictive hiring platforms, and performance‑scoring algorithms. Employers must provide a six‑point written notice—including the tool’s trade name, data categories, and contact information—whenever AEDT materially influences a decision. Additionally, the law amends Connecticut’s anti‑discrimination framework, making it clear that reliance on AI does not shield companies from bias claims, and retains a private right of action for affected workers.

Practically, firms should launch a comprehensive AI inventory, flagging any system that informs personnel decisions and assessing its material impact. Vendor contracts need revision to capture disclosure obligations, and HR and legal teams must be trained on plain‑language notice templates. Early compliance not only mitigates the risk of AG enforcement—treated as unfair or deceptive trade practices—but also positions companies as responsible stewards of algorithmic governance, a competitive advantage as AI adoption accelerates across industries.