What happened
On August 30, 2026, the California Legislature gave final passage to SB 813 (authored by Sen. Jerry McNerney, sponsored by Fathom), which directs the Government Operations Agency to create a framework for designating Independent Verification Organizations (IVOs) — qualified third-party auditors empowered to assess AI safety claims and measure risk in AI systems and models. The bill adds a new chapter to the California Government Code, requiring the Agency to develop IVO application/designation criteria, suspension/termination procedures, and stakeholder working groups by January 1, 2028. The bill now goes to Governor Newsom, who has until the end of September 2026 to sign or veto.
Why it matters
This would be the first state framework in the US for independent, third-party verification of AI safety claims, building on California's SB 53 (frontier developer safety-framework/incident-reporting law). It moves AI safety assessment away from self-reporting by developers toward accountable third-party auditors, a structural shift that could become a template followed by other states given the current federal regulatory vacuum on frontier AI safety.
Action needed
Monitor for Governor Newsom's signature/veto decision (due by end of September 2026); frontier AI developers operating in California should track IVO designation criteria once published by the Government Operations Agency (due by Jan 1, 2028).