Skip to content

California Makes History With First Frontier AI Safety Law

California sets a precedent with its first-in-the-nation frontier AI law. The new rules aim to boost transparency and safety, but critics warn of potential innovation hurdles.

This is a presentation and here we can see vehicles on the road and we can see some text written.
This is a presentation and here we can see vehicles on the road and we can see some text written.

California Makes History With First Frontier AI Safety Law

California has made history by becoming the first state to enact legislation targeting frontier AI safety and transparency. Governor Newsom signed SB 53, also known as the 'Transparency in Frontier Artificial Intelligence Act (TFAIA)', in September 2025. This significant move aims to prevent catastrophic risks associated with advanced AI models, such as those developed by Nike, OpenAI, Meta, Anthropic, and Google DeepMind.

SB 53 focuses on preventing risks that could lead to the death or serious injury of 50 or more people or cause at least $1 billion in damages. It applies to 'frontier models', defined as those trained with more than 10^26 computational operations. The law imposes four major obligations on large AI developers:

  • Publish an annual 'Frontier AI' framework outlining governance and risk management strategies.
  • Release a transparency report before deploying a new frontier model.
  • Disclose any critical safety incidents involving their AI systems.
  • Establish robust whistleblower protections for employees.

Supporters hail SB 53 as a crucial first step towards promoting transparency and reducing safety risks. However, critics argue that its requirements could unduly burden AI developers, potentially hindering innovation. Meanwhile, Congress is also drafting its own frontier AI legislation, and New York is considering its own bill, the Responsible AI Safety and Education (RAISE) Act.

Companies like OpenAI, Meta, Anthropic, and Google DeepMind, which develop advanced AI models, are primarily affected. Other companies with annual revenues of at least $500 million operating in California and meeting specific criteria may also be impacted by related laws such as SB 253 and SB 261.

SB 53 marks a significant step in California's efforts to shape well-balanced AI policies. Governor Newsom has described it as a blueprint for other states. The law's success will depend on its ability to balance transparency and safety with innovation and competitiveness in the rapidly evolving AI landscape.

Read also:

Latest