Pelosi slams California bill that would regulate artificial intelligence: ‘misinformed’

Representative Nancy Pelosi has come out against a high-profile California state bill that would regulate Big Tech’s fast-growing artificial intelligence technology — describing the bill as “well-intentioned but misinformed.”

California’s SB 1047 would set safety standards for AI models that cost more than $100 million to train and mandate pre-release safety testing for so-called “frontier” AI models, among other requirements. It is opposed by the tech industry, including Meta, Google and venture firm Andreesen Horowitz.

Pelosi (D-Calif.) — who has come under fire for owning a lucrative stock portfolio that includes AI chip supplier Nvidia, software firm Databricks and other big tech firms while still in office — argued that the law can stifle innovation.

“While we want California to lead in AI in a way that protects consumers, data, intellectual property and more, SB 1047 does more harm than good in that pursuit,” Pelosi said in a lengthy statement.

Pelosi noted that several California Democratic lawmakers, including Reps. Zoe Lofgren, Anna Eshoo and Ro Khanna and Gov. Gavin Newsom, have expressed concerns about the legislation or opposed it outright.

She also cited the view of Stanford University artificial intelligence researcher Fei-Fei Li, who has warned that the bill “would have significant unintended consequences that would stifle innovation and damage the US AI ecosystem,” according to Pelosi’s statement.

The bill was introduced by California state Sen. Scott Weiner, who is widely expected to run against Pelosi’s daughter, Christine, to take her seat in the US House when the former speaker leaves office.

Weiner said he “respectfully and strongly” disagreed with Pelosi.

“Innovation and security are not mutually exclusive, and I reject the false claim that in order to innovate, we must leave security solely in the hands of technology companies and venture capitalists,” Weiner said.

The legislation is a key fight in the national debate over how the government should regulate powerful AI models built by Google and Sam Altman’s OpenAI.

Critics warn that advanced artificial intelligence systems could pose a huge risk to society – from spreading misinformation to destroying humanity – without proper safeguards.

SB 1047 will receive a vote in the California state assembly by the end of August.

Opponents of the bill claim the legislation will only hurt the US in its bid to lead the advanced AI race.

In June, Meta AI chief Yann LeCun argued that the bill’s approach to AI regulation “will kill innovation.”

In an effort to assuage industry concerns, California lawmakers made changes to the bill earlier this month, such as dropping plans to create a government agency called the Frontier Model Division (FMD) to oversee regulation. of security.

Lawmakers also removed a clause that would have allowed the California state attorney general to directly sue AI firms if they were found to be negligent in their security practices.

Anthropic, the AI ​​giant backed by Google and Amazon, had pushed for this change. After the changes were announced, Anthropic told TechCrunch that it was still reviewing the amended bill and that not all of its proposals had been approved.

However, the amended version of the bill has not satisfied all opponents.

“The redactions are window dressing,” Andreesen Horowitz partner Martin Casado wrote in X. “They do not address the real issues or criticisms of the bill.”