California signed SB 813 and AB 1405, described by policy trackers as the first U.S. AI audit laws. They require certified auditors for AI used in hiring, insurance, and other critical decisions, with obligations phasing in by 2029.
The bills sit on top of a 2026 session in which more than 20 AI measures reached the governor’s desk. While Washington debates a federal duty of care and preemption, Sacramento is building an audit market the way it once built a privacy-compliance market after CCPA.
Employers and insurers that already deploy scoring models now have a calendar: find certified auditors, document systems, and treat model changes as audit events. Vendors that sell hiring or underwriting AI will be pulled into the same paperwork.
National labs still focus on catastrophic frontier risk. California’s first move is narrower and more operational—audits for decisions that already affect jobs and premiums—setting a template other states can copy before Congress finishes its draft.
Key takeaway The first U.S. AI audit mandate is state law, aimed at hiring and insurance systems, with a 2029 compliance horizon.
Photo: San Francisco via Unsplash. Sources: Enterprise DNA policy brief, September 13, 2026; California legislative coverage.
