Governor Gavin Newsom signed two bills that set up a system for independent organizations to audit AI systems and models operating in California.

 

What happened

Newsom signed SB 813 and AB 1405 on Wednesday, 9 September 2026. SB 813 creates a framework for independent verification organizations to assess AI systems and models. AB 1405 establishes a state registry for AI auditors and sets standards for their independence, transparency, and integrity. The laws set up a system for evaluating whether AI models comply with state law, as AI becomes more embedded in government services, critical infrastructure, and other sectors.

 

The backstory

California already has a set of AI regulations. The Transparency in Frontier Artificial Intelligence Act (SB 53) requires frontier AI developers to disclose their safety practices and report certain incidents. The California AI Transparency Act (SB 942) sets disclosure rules for AI-generated content, requiring large generative AI providers to label AI-created images, video, and audio. Furthermore, new automated decision-making technology (ADMT) regulations under the California Consumer Privacy Act govern how businesses use AI systems to make significant decisions about consumers. SB 813 and AB 1405 add another layer to that existing regulatory structure.

Read also: California Bill would bar AI from replacing licensed clinical judgment

 

Going deeper

California's move follows a similar step by Illinois. In July, Gov. JB Pritzker signed the Artificial Intelligence Safety Measures Act, which requires annual independent third-party audits for major frontier AI developers. California and Illinois are now both requiring outside verification of AI systems, though through different legal mechanisms, Illinois mandates the audits directly, while California builds a registry and framework to standardize who can perform them.

 

What was said

Newsom said in a statement, "The concerns raised in recent incidents reinforce what California has long recognized: artificial intelligence holds extraordinary promise, but it must be developed and deployed with meaningful safeguards to protect the public."

He also said the pace of AI development requires federal involvement, "California has taken nation-leading action to advance AI safety, transparency, accountability, and responsible innovation, but the scale and potential consequences of this technology demand sustained action from every level of government." He added, "The federal government must step forward with robust, national regulations that match the urgency of this moment."

 

In the know

Third-party AI audits involve an outside organization, separate from the company that built or deployed the AI system, reviewing that system for safety, bias, or legal compliance. A registry of approved auditors, like the one AB 1405 creates, gives regulators and the public a way to confirm that an auditor meets baseline standards for independence and transparency before its findings can be trusted.

 

Why it matters

California isn't just legislating AI, it's actively building AI into how the state itself operates. The state recently launched AskCA, an AI assistant for residents navigating government services, has given agencies access to Anthropic's Claude tools, and has state employees testing an internally built assistant called Poppy. It has also tightened oversight of AI vendors seeking state contracts, requiring them to address bias, civil rights violations, and harmful content. This means California is regulating a technology it's also adopting. A verified, registered pool of auditors matters because it gives the state, and other buyers, a way to check the same AI vendors and tools it's bringing into public services, rather than taking vendor safety claims on faith.

 

The bottom line

States are stepping in to police AI safety in the absence of federal rules. Newsom's own call for national regulation suggests even California sees its framework as a stopgap rather than a permanent solution. For now, AI developers and vendors working with California agencies will need to work with a formal audit and registry system built to check their work.

 

FAQs

What is a frontier AI model?

It refers to the most advanced, large-scale AI systems trained with high computing power, which is why regulators single them out for extra scrutiny.

 

Why are states regulating AI instead of the federal government?

In the absence of federal AI legislation, individual states have been passing their own laws to address safety, transparency, and accountability concerns.

 

What does an AI auditor check for?

Depending on the law, auditors evaluate an AI system for things like safety risks, bias, and compliance with applicable regulations.