(Bloomberg) — California Gov. Gavin Newsom vetoed what would have become one of the most comprehensive policies governing the safety of artificial intelligence in the United States.
Most read on Bloomberg
The bill would have been one of the first to hold AI developers accountable for any serious harm caused by their technologies. It drew sharp criticism from prominent Democrats and major tech companies, including ChatGPT creator OpenAI and venture capital firm Andreessen Horowitz, who warned it could block innovation in the State.
Newsom called the legislation “well-intentioned” but said in a statement that it would have applied “high standards to even the most basic functions.” Regulation should be based on “empirical and scientific evidence,” he said, pointing to his own executive order on AI and other bills he has signed that regulate the technology around known risks such as deepfakes.
The debate around California’s bill SB 1047 highlights the challenge facing lawmakers around the world in controlling AI risks while supporting the emerging technology. U.S. policymakers have yet to pass comprehensive legislation around the technology since the release of ChatGPT two years ago sparked a global boom in generative AI.
California Democratic Sen. Scott Wiener, who introduced the bill, called Newsom’s veto “a setback for everyone who believes in corporate oversight.” In a statement published on X, Wiener said: “As a result, we are all less safe. »
“Reasonable care”
SB 1047 would have required companies developing powerful AI models to reasonably ensure that their technologies do not cause “serious harm,” such as mass casualties or property damage exceeding $500 million. Companies should have taken specific precautions, including maintaining a kill switch that could disable their technology. The AI models would also have undergone third-party testing to ensure they minimized serious risks.
The bill also would have created whistleblower protections for employees of AI companies who want to share their security concerns. Companies that failed to comply with the bill could have been sued by the California Attorney General.
Supporters of the legislation said it would have created common-sense legal standards. But venture capitalists, startup executives and companies like OpenAI have warned it will slow innovation and drive AI companies out of the state.
“The AI revolution is only just beginning, and California’s unique status as a global leader in AI is fueling the state’s economic dynamism,” wrote Jason Kwon, chief strategy officer at OpenAI, in a letter opposing the legislation last month. “SB 1047 would threaten that growth, slow the pace of innovation, and drive California’s world-class engineers and entrepreneurs to leave the state in search of better opportunities elsewhere.”
Lawmakers oppose it
Lawmakers including former House Speaker Nancy Pelosi, Rep. Ro Khanna and San Francisco Mayor London Breed also expressed opposition, echoing concerns from the tech industry that the bill could hinder California’s leadership in AI innovation. Newsom recently said he was concerned the bill would have a “chilling effect” on AI development.
The bill had gained support from some notable names in the tech sector late last month, in the days before it was passed by the California legislature. Elon Musk unexpectedly expressed his support, although he said it was “a difficult decision and one that will upset some people.” OpenAI rival Anthropic, which has a reputation for being focused on security, said the bill’s “benefits likely outweigh its costs,” although the company said some aspects remained “concerning or ambiguous for us.
Wiener had defended the bill, emphasizing that its provisions only applied to companies that spend more than $100 million on training large models or $10 million on fine-tuning models, limits that would exempt the most small startups. The lawmaker also noted that Congress has historically been slow to regulate the technology itself.
In announcing his veto, Newsom said he would consult with outside experts, including AI specialist and entrepreneur Fei-Fei Li, to “develop workable safeguards” on the technology and continue working with the legislature. the State on the subject.
The governor also signed a bill, SB 896, on Sunday that regulates how state agencies use AI.
(Updates with Wiener’s statement in fifth paragraph)
Most read from Bloomberg Businessweek
©2024 Bloomberg LP