Anthropic Endorses California's AI Safety Bill SB 53

Anthropic has endorsed SB 53, a California bill that aims to regulate powerful AI systems developed by major AI companies. Announced in a press release, the bill, introduced by Senator Scott Wiener, focuses on transparency and safety protocols for large AI developers.

SB 53 requires companies to develop and publish safety frameworks, release public transparency reports, and report critical safety incidents within 15 days. It also includes whistleblower protections and mandates public accountability for commitments made in safety frameworks.

The bill targets large AI companies, such as OpenAI, Google, and Anthropic, while providing exemptions for smaller companies. This approach aims to ensure that powerful AI advancements are managed responsibly without stifling innovation among smaller developers.

Anthropic's endorsement of SB 53 reflects its commitment to AI safety and transparency, aligning with its existing practices of publishing safety reports and system cards. The bill is seen as a step towards establishing a regulatory framework that could influence AI governance beyond California.