[ Answers · AI ]

What is California SB 53?

By Ben Tabas, Intrica Research · Updated · US

Answer

California SB 53, the Transparency in Frontier Artificial Intelligence Act, sets safety and transparency rules for developers of the largest AI models. Governor Newsom signed it on Sept. 29, 2025, and most duties took effect Jan. 1, 2026, per a July 9, 2026 National Law Review summary. Some reporting provisions start Jan. 1, 2027.

Who it covers

  • A frontier model is a foundation model trained with more than 10^26 integer or floating point operations, per the National Law Review on July 9, 2026.
  • A frontier developer is anyone who has trained, or started training, a frontier model, per the same summary.
  • A large frontier developer is a frontier developer whose annual gross revenue, with affiliates, exceeded $500 million in the prior calendar year, per the same summary.

What it requires

  • Every frontier developer must publish a transparency report at or before deploying a new or substantially modified frontier model, per the National Law Review on July 9, 2026.
  • Large frontier developers must publish a frontier AI framework, review it at least once a year and update it within 30 days of a material change, per the same summary.
  • Large frontier developers must send summaries of catastrophic risk assessments to the Office of Emergency Services every three months unless it sets another schedule, per Clifford Chance on Oct. 24, 2025.
  • Critical safety incidents go to the Office of Emergency Services within 15 days, or within 24 hours when there is an imminent risk of death or serious injury, per the National Law Review on July 9, 2026.
  • Employees who report safety risks are protected from retaliation, per the Governor's office on Sept. 29, 2025.
  • The California Attorney General can seek civil penalties of up to $1 million per violation, per Clifford Chance on Oct. 24, 2025.

What starts Jan. 1, 2027

  • The Office of Emergency Services begins publishing anonymized yearly summaries of critical safety incidents, per the National Law Review on July 9, 2026.
  • The Attorney General begins publishing anonymized yearly reports on whistleblower disclosures, per the same summary.
  • A consortium delivers its report on CalCompute, a proposed state public cloud computing cluster, per the same summary.

Status as of Oct. 28, 2026

The National Law Review described the law's obligations as in effect on July 9, 2026, with some provisions starting Jan. 1, 2027.

The California Department of Technology must review the law's frontier developer definitions each year and can recommend updates, per the same summary. Check those reviews before relying on the thresholds.

See whether ChatGPT and Google AI cite your company when buyers ask how AI vendors handle model safety.

Run the free Answer ShareBook a call

Sources

  1. Governor Newsom signs SB 53, advancing California's world-leading artificial intelligence industry · Office of the Governor of California, Sept. 29, 2025Supports: The signing date, the bill's name and its main provisions, including whistleblower protection.
  2. California's SB 53: understanding obligations in effect now and in 2027 · National Law Review, July 9, 2026Supports: Effective dates, coverage thresholds, reporting deadlines and the provisions that start Jan. 1, 2027.
  3. SB 53: California sets standards for AI transparency · Clifford Chance, Oct. 24, 2025Supports: The quarterly catastrophic risk summaries, the $1 million penalty and Attorney General enforcement.

Related questions

General information, not legal advice.

Intrica helps AI companies get found in Google and named in AI answers. SEO and AI search for AI companies →