California Lawmakers Pass AI Safety Bill SB 53, But Governor Newsom's Veto Decision Looms Large

California Lawmakers Pass AI Safety Bill SB 53, but Newsom's Veto Looms Large In a major development for the tech industry, California lawmakers passed bill SB 53 on Saturday morning, requiring large AI labs to be transparent about their safety protocols and creating whistleblower protections for employees. The legislation now heads to Governor Gavin Newsom's desk, where it may face a veto. The bill, authored by state senator Scott Wiener, was given final approval in the state senate after months of debate. According to Wiener, SB 53 sets new transparency requirements on large companies, creates a public cloud to expand compute access (CalCompute), and establishes whistleblower protections for employees at AI labs. "We're taking an important step forward in ensuring that AI is developed and deployed safely," Wiener said in a statement. "This bill will help prevent the kind of catastrophic failures we've seen with other technologies." SB 53 was influenced by recommendations from a policy panel of AI experts convened by Newsom last year, after he vetoed a more expansive safety bill also authored by Wiener. The new legislation has been amended to address concerns raised by critics. According to Politico, the amendments were made to ensure that companies are held accountable for their actions, regardless of whether they deploy large models in high-risk environments or use sensitive data. Newsom's office has not commented on SB 53, but his previous veto of Wiener's safety bill highlighted the governor's concerns about applying stringent standards to large models. At the time, Newsom acknowledged the importance of protecting the public from real threats posed by AI technology. The implications of SB 53 are far-reaching and could have significant impacts on the development and deployment of AI in California. Large companies like Google, Facebook, and Microsoft will be required to disclose their safety protocols and ensure that employees can report any concerns without fear of retaliation. While some experts welcome the new legislation as a step forward for AI safety, others argue that it does not go far enough. "This bill is a good start, but we need more comprehensive regulations to address the risks associated with large language models," said Dr. Timnit Gebru, a leading expert on AI ethics and bias. The current status of SB 53 remains uncertain, as Newsom has yet to indicate whether he will sign or veto the legislation. If signed into law, SB 53 would take effect in January 2024. As the tech industry continues to grapple with the challenges posed by AI, California lawmakers have taken a crucial step forward in ensuring that this technology is developed and deployed safely. The fate of SB 53 now rests in Newsom's hands, as he weighs the importance of protecting public safety against concerns about over-regulation. *Reporting by Techcrunch.*