California Governor Gavin Newsom has signed the groundbreaking AI Safety Bill SB 53, making California the first U.S. state to enforce strict AI transparency rules on major tech firms.

The AI Safety Bill SB 53, passed by lawmakers just two weeks ago, requires large AI companies — including OpenAI, Meta, Anthropic, and Google DeepMind — to disclose their safety protocols and report critical incidents involving AI systems.

Under the new law, AI labs must notify California’s Office of Emergency Services about potential risks such as cyberattacks, deceptive AI behaviors, or crimes committed without direct human oversight. The bill also guarantees whistleblower protections for employees who expose safety risks.

While Anthropic supported the bill, tech giants like Meta and OpenAI strongly opposed it. OpenAI even published an open letter urging Governor Newsom not to sign the legislation, citing concerns about overregulation and its impact on innovation.

However, Newsom emphasized that AI safety and innovation must coexist:

“California is proving we can protect communities while ensuring AI continues to thrive,” Newsom said. “This first-in-the-nation legislation sets the standard for responsible AI governance.”

The passage of SB 53 could spark similar legislation across the U.S., as other states like New York are already considering comparable measures.

Meanwhile, another proposal, SB 243, which seeks to regulate AI companion chatbots and hold developers legally accountable for harmful interactions, is awaiting the governor’s decision.

This marks Senator Scott Wiener’s second attempt to pass AI safety laws after his earlier bill, SB 1047, was vetoed last year amid intense tech industry backlash. By working closely with AI companies this time, Wiener aimed to strike a balance between public safety and technological progress.

As the AI Safety Bill SB 53 takes effect, California positions itself as a leader in shaping the future of safe and transparent AI development in the U.S.