Governor Gavin Newsom today vetoed a major AI safety bill, saying, “As long as AI is deployed in large systems, even the most basic functions will be subject to strict standards.” .
“I do not believe this is the best approach to protecting the public from the real threats posed by technology,” Newsom said in a statement announcing his veto.
SB 1047 would have required AI developers in the state to take security measures before training their models. OpenAI had warned of the bill’s impact on industry growth, but former House Speaker Nancy Pelosi (D-Calif.) opposed the bill.
Celebrities such as Jane Fonda, Shonda Rhimes and Pedro Pascal have urged Newsom to sign the bill, and its lead author, state Sen. Scott Weiner, said the bill would “result in the largest AI companies already “It’s a light-touch, common-sense measure that codifies the commitments we’ve made.” It was created voluntarily. ”
“This veto is a missed opportunity for California to once again lead on innovative technology regulation, much like it has done on data privacy and net neutrality,” Weiner said in a statement. As a result, we are all less safe.”
Newsom said he will continue to work with “Congress, federal partners and technology.”
Experts, ethicists, and academia can identify appropriate paths forward, including through legislation and regulation. ”
“Safety protocols must be adopted,” Newsom wrote. “Aggressive guardrails must be implemented, and serious consequences for malicious actors must be clear and enforceable. We do not agree that we must settle for solutions that are not based on an empirical trajectory analysis of systems and capabilities. Ultimately, the framework for effectively regulating Al depends on the technology itself. We need to keep pace.”
Earlier this month, Newsom signed a series of other AI bills. These include bills that would give performers more rights over their image and likeness, as well as legislation aimed at curbing the use of deepfakes in election campaigns.
