California adopts new AI laws requiring independent audits, assessments
The laws create independent audit standards and a state registry as AI companies face growing pressure to prove their systems are safe.
- On Wednesday, California Governor Gavin Newsom signed SB 813 and AB 1405, establishing the nation's first framework for independent third-party AI audits and a state registry for auditors.
- Recent incidents involving autonomous agents exploiting software flaws between May and July exposed critical gaps in the state's 2025 Transparency in Frontier Artificial Intelligence Act, known as SB 53.
- Industry leaders OpenAI and Anthropic publicly backed the new legislation, aiming to establish a 'national baseline' for safety as both firms prepare for initial public offerings amid increased scrutiny.
- Congress faces a December adjournment deadline as OpenAI Chief Global Affairs Officer Chris Lehane urges lawmakers to adopt 'mandatory, capability-based national AI safety rules' before the session ends.
- Prediction markets assign an 86% probability to Anthropic reaching a $2.0 trillion valuation by year's end, reflecting market confidence as officials emphasize that national standards must align with international safety protocols.
38 Articles
38 Articles
Artificial Intelligence Regulation in California: Discover the two laws Newsom signed to audit and control its progress.
California Certifies the AI Watchdogs: New Laws Create First U.S. Framework for Independent Model Testing
Governor Newsom signed SB 813 and AB 1405 this week, creating the nation's first framework for state-certified Independent Verification Organizations and an auditor registry. The laws aim to replace company self-assessments with qualified third-party testing of AI safety risks. Early examples show high costs and potential conflicts, yet the voluntary system marks a major accountability step.
This call comes amid growing fears about the impact of AI.
OpenAI urged the U.S. Congress to pass mandatory security laws for artificial intelligence by December. Learn the details.
OpenAI urges the U.S. Congress to establish mandatory standards for AI, highlighting that voluntary agreements are not enough to mitigate risks.The company seeks a federal framework with common testing standards and strict cybersecurity, amid fears about the uncontrolled impact of AI.
Coverage Details
Bias Distribution
- 53% of the sources lean Right
Factuality
To view factuality data please Upgrade to Premium
