Bill > SB53

This bill establishes the Transparency in Frontier Artificial Intelligence Act (TFAIA), which creates comprehensive regulations for large frontier developers of advanced AI models. The legislation requires these developers to create and publish a detailed "frontier AI framework" that outlines how they will assess and mitigate potential catastrophic risks from their AI models, such as risks involving potential weapon creation, cyberattacks, or loss of model control. Large frontier developers (defined as those with annual revenues over $500 million) must report critical safety incidents to the Office of Emergency Services within 15 days, conduct regular risk assessments, and establish internal whistleblower mechanisms that allow employees to anonymously report potential dangers. The bill also creates a new consortium to develop "CalCompute," a public cloud computing platform to advance safe and ethical AI research, and establishes strong whistleblower protections for employees who report potential AI-related risks. The legislation aims to increase transparency in AI development, protect public safety, and ensure that developers are taking responsible precautions with increasingly powerful artificial intelligence technologies. Critically, the bill includes provisions for civil penalties up to $1 million for non-compliance and includes safeguards to protect trade secrets and national security while promoting public accountability.

Budget and Finance, Business and Industry, Government Affairs, Justice

https://leginfo.legislature.ca.gov/faces/billStatusClient.xhtml?bill_id=202520260SB53