California Gov. Gavin Newsom signed two artificial intelligence safety bills Wednesday, creating new state standards for independent groups that evaluate AI systems and their potential risks.

The measures, Senate Bill 813 and Assembly Bill 1405, establish what California describes as the nation’s first framework for independent AI safety assessments. SB 813 creates a process for designating qualified organizations that can independently assess AI systems for compliance with state law, while AB 1405 establishes a registry and standards for AI auditors.

The new framework is intended to give outside evaluators standards for independence, transparency, and technical qualifications when reviewing advanced AI systems. Under SB 813, California’s Government Operations Agency is required to establish qualification criteria for independent verification organizations by 2028. The program is voluntary.

OpenAI endorsed the legislation along with two other California AI bills, saying independent safety assessments should become part of a broader system of mandatory AI safety requirements. The company said SB 813 would establish a process for designating qualified independent organizations capable of assessing AI risks, while AB 1405 would set standards for AI auditors.

Anthropic also supported the broader effort as concerns about AI safety have intensified. The company has faced renewed scrutiny following incidents involving AI systems accessing external computer systems during testing.

The legislation was signed amid heightened attention surrounding Jacob Coxon, a former researcher who worked at both Anthropic and OpenAI before resigning from Anthropic and publicly warning about the risks he believes could arise from increasingly capable AI systems. His departure and comments have fueled renewed debate among researchers and lawmakers over how advanced AI should be evaluated and regulated.

Newsom said California's framework is intended to increase transparency and accountability while the technology continues to develop. The governor has also called for federal AI regulations, arguing that national standards are needed alongside state-level safeguards.