California’s legislators pass AI Safety Bill Sb 53 – but Newsom could still veto

The California State Senate gave final approval Early on Saturday morning on an important AI security bill that sets new transparency requirements in large companies.

As described by his author, State Senator Scott Wiener, Sb 53 “Requires large AI laboratories to be transparent about their security protocols. [employees] At AI Labs & creates a public cloud to extend calculation access (Calcompute). ”

The bill is now going to California Governor Gavin Newsom to sign or veto. It has not publicly commented on SB 53, but last year, veto on a more expansive security bill that was also compiled by Wiener, while signing closer legislation targeting issues such as Deepfakes.

At that time, Newsom recognized the importance of “protecting the public from real threats that this technology put”, but criticized Wiener’s previous Wiener bill to large models, regardless of whether they “developed in high -risk environments” [involved] critical decision -making or use of sensitive data. ”

Wiener said the new bill was influenced by recommendations by a group of political experts that convened the Newsom after its veto.

Politico also reports This SB 53 has recently been modified so that companies will develop AI models “Frontier”, while having less than $ 500 million in annual revenue will only need to disclose high level security details, while companies that need to provide more detailed reports.

The bill has been opposed by several Silicon Valley companies, VC companies and pressure groups. In A recent letter to NewsomOpenai did not specifically mention SB 53, but argued that in order to avoid “overlapping and inconsistencies”, companies should be considered complying with state security rules, provided they meet federal or European standards.

And head of AI policy and Andreessen Horowitz’s leader Recently claimed that “many of today’s AI government accounts – such as proposals in California and New York – danger” crossing a line violating the constitutional boundaries of how states can regulate transnational trade.

The A16Z co -founders had previously highlighted the technological regulation as one of the factors that led them to support Donald Trump’s offer for a second term. Trump’s administration and its allies then demanded a 10 -year ban on the AI ​​State Regulation.

In the meantime, it has come out in favor of SB 53.

“We have long said that we would prefer a federal standard,” said human co -founder Jack Clark In a post. “But, in the absence of this, this creates a compact plan for the AI ​​governance that cannot be ignored.”