The wild west days of little to no guardrails on workplace AI are ending – not because of any federal regulations but because states are closing the frontier. Several states passed consequential workplace AI laws in 2026 which are either already in effect or will be starting in 2027.
These laws largely fall into three buckets:
-
Discrimination: Laws targeting AI-related discrimination, requiring audits and regular anti-bias testing and recordkeeping (Colorado, California, Texas, Illinois, New York City).
-
AI in employment decisions: Laws restricting the use of AI in employment decisions, with requirements for notice to applicants/employees and humans in the loop (Connecticut, and soon California).
-
AI-related job impacts: Laws requiring companies to disclose AI-related job impacts data to the state (Connecticut, New York, California).
Expect several new laws in these areas to emerge before the end of 2026 and into 2027. Even blue states have been hesitant to overregulate AI, but the dam has finally burst. That flow will only pick up pace once AI-related job impacts start hitting home.
Unfortunately for multi-jurisdictional employers, the lack of federal action makes a state patchwork inevitable. We will be watching to see how our members set their compliance approaches.
