California Mandates Transparency for Large AI Models

Gavin Newsom, the Governor of California, has signed the Transparency in Frontier Artificial Intelligence Act (TFAIA), also known as SB 53, into law. This legislation mandates increased transparency and reporting requirements for developers of large AI models, known as 'frontier models.' The law will take effect on January 1, 2026.

The TFAIA requires 'frontier developers' to publish a 'frontier AI framework' on their websites, detailing technical and organizational protocols to manage and mitigate catastrophic risks associated with their AI models. These frameworks must be reviewed and updated annually. Additionally, developers must publish transparency reports when deploying new or substantially modified models, outlining the model's intended uses and any restrictions.

The law also mandates that developers report 'critical safety incidents' to the California Office of Emergency Services within specific timeframes, depending on the severity of the incident. Whistleblower protections are included, allowing employees to report potential dangers or violations anonymously.

This legislation is part of a broader effort by the California Legislature to regulate AI development and ensure public safety, aligning with similar initiatives in other jurisdictions such as the EU AI Act.