Published: 22 Aug 2024 03:42 GMT
Several technology companies developing artificial intelligence have opposed the project, arguing that it threatens growth and innovation in the state.
State lawmakers are set to vote this week on a bill that would broadly regulate the development of artificial intelligence (AI) in the US state of California, despite opposition from several Silicon Valley giants.
The proposal, known as SB 1047 and introduced by Democratic Senator Scott Wiener, would require safety testing for many of the most advanced AI models—specifically, those that cost more than $100 million to develop. 100 million dollars or that require a certain computing power, depending on collect Reuters.
If approved by the California Congress, AI software developers operating in the state will have to submit detailed documentation methods to disable models AI in case they fail, in addition to hiring external auditors to evaluate their security practices.
The bill would also give state Attorney General Rob Bonta the power to sue developers who break the rules, especially in cases where AI poses ongoing threats and endanger public systems like the power grid.
SB 1047 has already been approved in the Senate by 32 votes in favor and one against. It also received the green light from the appropriations committee of the state assembly, so it will be submitted to a vote in the plenary session and, if approved before August 31, it will go to the state governor, Gavin Newsom, to be signed.
Brilliant minds will leave California
While several tech companies have called for stronger security guardrails, citing the risk that software could one day bypass human intervention and lead to cyberattacks, they have also largely opposed SB 1047.
OpenAIdeveloper of ChatGPT, broke its silence on the bill on Tuesday, officially expressing its opposition in a letter to Wiener and Newsom. The AI giant argued that SB 1047 would stifle innovation and push talent out of California.
“The AI revolution has only just begun, and California’s unique status as a global leader in AI is fueling the state’s economic dynamism,” said OpenAI Chief Strategy Officer Jason Kwon at the cartaobtained by TechCrunch. “SB 1047 would threaten that growth“would slow the pace of innovation and drive California’s world-class engineers and entrepreneurs to leave the state in search of greater opportunities elsewhere,” he added.
The company joined the broad local pressure against the law, shared by trade groups representing Google y Meta*the investment company Andreessen Horowitzand California Reps. Nancy Pelosi and Zoe Lofgren, who argue that given the risks, America’s AI advantage should be protected with a set of federal, rather than state, policies that can provide clarity and certainty to AI labs and developers across the country.
*Classified in Russia as an extremist organization, whose social networks are banned in its territory.
