A new bill aimed at regulating AI companies has been tabled in California after a similar attempt failed to pass through last year

A new bill aimed at regulating AI companies has been tabled in California after a similar attempt failed to pass through last year. Introduced by Senator Scott Wiener, SB 53 takes a fresh stab at managing the growing power of artificial intelligence. It comes on the heels of SB 1047, a broader AI safety bill that Governor Gavin Newsom vetoed in September 2024. That earlier effort stirred up a storm, with tech leaders arguing it would stifle innovation while supporters said it was needed to prevent disasters caused by massive AI systems.

This time, SB 53 narrows the focus. It zeroes in on protecting whistleblowers at top AI labs like OpenAI or Anthropic. These are employees who might spot serious dangers in their company’s AI projects. The bill defines those dangers as risks that could kill more than 100 people or cause over $1 billion in damages.

Workers could report their worries to California’s Attorney General, federal officials, or even coworkers without facing punishment from their bosses. Companies would also have to explain how they’re tackling the issues these whistleblowers raise. This push for transparency stands out, especially after cases like Suchir Balaji’s. He was an OpenAI employee who went public with concerns about the company’s work, sparking debate about the need for protections like these.

SB 53 doesn’t stop there. It also lays out plans for CalCompute, a public cloud computing system. Think of it as a shared supercomputer for researchers and startups. Big tech firms often hog the computing power needed for cutting-edge AI, leaving smaller players in the dust. CalCompute aims to change that by giving more people a shot at building AI that serves the public good. University of California reps and other experts would figure out how to set it up and who gets access.

Will SB 53 make it into law? It’s got a long road ahead through California’s legislature. The tech crowd might push back again, just like they did with SB 1047. Plus, the mood around AI safety seems to be shifting. Vice President J.D. Vance recently hinted that the US cares more about speeding up AI progress than slowing it down with rules.

Still, SB 53 tries to strike a balance — shielding workers and boosting research without the heavy-handed controls that sank its predecessor. Its fate could shape how California, and maybe the country, handles AI’s rise.