California AI Safety Bill Faces Final Hurdle
The proposed law demands rigorous safeguards for AI systems that pose the highest risks.
Jericho Casper
WASHINGTON, August 30, 2024 – A bill mandating safety testing for companies developing the most advanced generative AI systems is now heading to California Gov. Gavin Newsom's (D) desk for approval.
The AI safety bill, SB 1047, passed the California State Assembly with a 41-16 vote, and the State Senate by 29-9 in a floor vote on Thursday.
SB 1047 was designed to safeguard Californians by preventing AI misuse in instances of cyberattacks, weapons development, and automated crime. The bill would require developers of AI systems that cost over $100 million to train to test their models for potential harm and implement safety measures.
Despite SB 1047's focus on regulating only the largest AI systems, the bill has faced criticism from figures like former U.S. House Speaker Nancy Pelosi (D-Calif.), who argued that the potential for legal action from the state attorney general could stifle innovation in California, a hub for leading AI companies.
Supporters of the bill, however, have claimed it struck a crucial balance between innovation and public safety, positioning California as a leader in responsible AI governance.
State Senator Scott Wiener (D-San Francisco), who introduced the bill, said he amended the measure to reflect advice from leaders in the AI space, safety groups, academics, and developers like Amazon-backed Anthropic.
“It’s been an enormous amount of work and collaboration with people who like the bill and people who don’t like the bill,” Sen. Wiener said. “I’m really, really proud of our coalition.”
The bill has garnered the support of some of the most respected figures in artificial intelligence, including Geoffrey Hinton and Yoshua Bengio, who are often referred to as the 'Godfathers of AI’ for their pioneering work in deep learning.
It also received backing from Elon Musk, a long-time advocate for AI regulation.
In an X post on Monday, Musk stated, “All things considered, I think California should probably pass the SB 1047 AI safety bill. For over 20 years, I [have advocated] for AI regulation, just as we regulate any product/technology that [has] a potential risk to the public.”
With a deadline of Sept. 30, all eyes are on Gov. Newsom to see whether California takes the lead in setting a precedent for AI safety in the U.S.
