What happened
Concluding the 2026 signing period by 30 September, Gov. Newsom enacted 12 AI-related measures including AB 1609 (customer-service chatbot disclosure for large businesses), SB 947 (no discipline/termination based solely on automated decision systems without human corroboration, plus post-use notice), SB 503 (bias-assessment duties for AI clinical decision support), SB 1000/AB 2713 (AI Transparency Act amendments removing the 1M-user threshold and adding provenance-data inspection), AB 1979 (AI may not independently perform functions requiring a professional license), AB 1331/AB 1883 (workplace surveillance limits incl. neural data), SB 867/SB 1119 (companion-chatbot child protections) and SB 1111 (AI digital replica impersonation).
Why it matters
California — the home jurisdiction of most frontier labs — now has binding statutory obligations on chatbot disclosure, human-in-the-loop employment decisions, health-AI bias evaluation and AI transparency that apply to companies nationwide doing business in the state, converging with Colorado and Connecticut in the state-law patchwork while federal regulation stalls.
Action needed
Deployers with California reach should map effective dates (several are immediate or early-2027), update chatbot disclosure UX, employment ADM workflows to include human corroboration and notice, and stand up provenance/transparency tooling for covered providers.