California's amended AI legislation now in implementation phase — SB 189-style frontier-AI rules begin enforcement after the August 2025 delay window expires
California's amended AI legislation — pushed back to June 2026 implementation in the August 2025 amendment — is now in active enforcement window. The legislation covers frontier-model providers operating in California with thresholds for safety evaluations, incident reporting, and capability disclosures. For OpenAI, Anthropic, and the major California-headquartered labs, the compliance posture moves from preparation to operational.
The substantive piece is the federalism question. With Congress holding to a light-touch federal AI Action Plan and the Trump executive order's 30-day frontier voluntary-review framework, US AI regulation remains state-led for most binding rules. California, Texas, New York, and Colorado have each moved with their own frameworks; the procurement question for AI providers is which state-by-state disclosures they must publish on a rolling basis.
The international frame is that the EU and California are converging on similar disclosure obligations through different routes. The EU AI Act's August 2 GPAI deadline mandates publication; California's framework mandates safety-evaluation incident reporting. Both attach to deployment, not training. For the frontier lab compliance team, the practical posture is to ship one disclosure stack that satisfies both — which most labs are now doing.
Anecdotes AI — AI Regulations in 2025: US, EU, UK, Japan, China & More → · MetricStream — AI Regulation Trends 2026: Policies Across the US, UK & EU → · Mean.ceo — AI Regulation News | June, 2026 (Startup Edition) →