WASHINGTON/SAN FRANCISCO: California Governor Gavin Newsom vetoed a highly debated artificial intelligence safety bill on Sunday, responding to concerns from the tech industry that it could drive AI companies out of the state and stifle innovation.

Newsom criticized the bill for not considering whether an AI system is used in high-risk settings, involves critical decision-making, or handles sensitive data, arguing that it would impose “stringent standards” even on basic functions. He emphasized the need for “workable guardrails” by collaborating with experts in generative AI to establish a science-based trajectory analysis.

Additionally, he instructed state agencies to broaden their evaluations of risks associated with potential catastrophic events linked to AI usage.

Generative AI, which can produce text, images, and videos in response to open-ended prompts, has generated both excitement and concerns about job displacement, electoral interference, and possible catastrophic outcomes.

The bill’s author, Democratic State Senator Scott Wiener, argued that legislation is crucial to safeguard the public as AI advances could become unmanageable or uncontrollable. He expressed that the veto renders California less safe, leaving companies that are developing powerful technologies without binding regulations.

Wiener noted that voluntary industry commitments are not enforceable and rarely benefit the public. “We cannot afford to wait for a major catastrophe to occur before taking action to protect the public,” Newsom stated, although he disagreed with the notion of adopting a solution that lacks an empirical analysis of AI systems.

He affirmed his intention to collaborate with the legislature on AI legislation in the next session. This decision comes as U.S. Congress has stalled on legislative safeguards while the Biden administration is moving forward with regulatory proposals for AI oversight.

Newsom mentioned, “a California-only approach may well be warranted, especially absent federal action by Congress.”

The Chamber of Progress, a tech industry coalition, commended Newsom’s veto, stating that “the California tech economy has always thrived on competition and openness.”

The proposed bill would have required safety testing for many advanced AI models costing over $100 million to develop or those necessitating a specific level of computing power. AI developers in California would also have been mandated to define methods for disabling their models, effectively creating a “kill switch.” Moreover, it aimed to establish a state body to oversee the development of “Frontier Models,” which surpass the capabilities of current advanced models.

The bill faced strong opposition from various groups, including major tech companies like Alphabet’s Google, Microsoft-backed OpenAI, and Meta Platforms, which raised concerns about the proposal. Some Democrats in Congress, including Representative Nancy Pelosi, also opposed the bill. Proponents included Tesla CEO Elon Musk, who leads an AI firm called xAI. Amazon-backed Anthropic argued that the bill’s benefits likely outweighed its costs, though it flagged some concerning or unclear aspects.

Separately, Newsom signed legislation requiring the state to assess potential threats posed by generative AI to California’s critical infrastructure. The state is currently evaluating risks to energy infrastructure and plans to conduct similar assessments for water and communications infrastructure in the coming year.