California wrote rules for AI auditors while Congress kept talking

Senator Ted Cruz said he was working with Senate Majority Leader John Thune and Senator Amy Klobuchar on legislation covering catastrophic AI risks. Six House members proposed a separate bill in July that would require independent security audits of the most capable models. Senators Josh Hawley and Richard Blumenthal sent new letters seeking details about OpenAI's agent incidents.

Federal statements are not federal rules. Reuters reports that the Cruz, Thune and Klobuchar proposal has no public text or release date. Letters can force disclosure, but they cannot create testing standards or enforcement powers.

The debate also depends heavily on claims from the companies and researchers building these systems. Independent auditors need access to the models, test environments and incident records. Without that access, an auditor could certify a narrow demonstration and miss how the system behaves with real tools or failed safeguards.

The Senate proposal should define which models it covers, who accredits testers and what evidence can stop a deployment. The House bill should explain whether audit reports become public and how it protects security-sensitive details.

California's implementation will show whether auditors receive enough access to reproduce failures. Watch for published standards, conflict-of-interest rules and the first enforcement case. Those details will tell readers whether the law produces evidence or paperwork.