Source: AFP
A bill aimed at regulating powerful artificial intelligence models passed the California Legislature on Wednesday, despite outcry that it could kill the technology it wants to control.
“SB 1047 — the AI safety bill — just passed the floor of the Assembly. I’m proud of the diverse coalition behind this bill — a coalition that believes deeply in both innovation and safety” , said Democratic Sen. Scott Wiener of San Francisco, who sponsored the bill.
The bill faced a barrage of critics, including Democratic members of the US Congress, who argued that threats of punitive measures against developers in a nascent field would stifle innovation.
However, it won the reluctant support of Elon Musk, who argued that AI’s risk to the public justifies regulation.
Read also
Judicial reforms strain Mexico-US ties, spook investors
“This is a tough call and it’s going to upset some people,” he said Monday as he declared his support for X.
The bill, called the Safe and Secure Innovation for Frontier Artificial Intelligence Models Act, now goes to California Gavin Newsom’s desk for signature, though his position on the proposal is not known.
Newsom has until September 30 to sign the bill or veto it.
Dan Hendrycks, director of the Center for Artificial Intelligence Security, said the bill provides “a workable path forward” to enforce “safeguards to mitigate critical AI risk.”
The bill requires developers of large AI models to take precautions such as testing before deployment, simulating hacker attacks, installing cyber security safeguards and providing whistleblower protections.
In order to ensure passage of the legislation in Silicon Valley’s hometown, lawmakers made several changes to the bill, including replacing criminal penalties for violations with civil penalties, such as fines.
Read also
UK Starmer warns October 30 Budget will be ‘painful’
However, opposition to the bill remained, including from some powerful national figures.
“The view of many of us in Congress is that SB 1047 is well-intentioned but ill-informed,” influential Rep. Nancy Pelosi of California said last week, noting that top party members have shared their concerns with Wiener.
But Wiener argues that AI safety and innovation are not mutually exclusive, and that the amendments to the bill have addressed some of the concerns of critics.
OpenAI, the creator of ChatGPT, also opposed the bill, saying it would prefer national rules over a chaotic patchwork of AI regulations across the 50 US states.
At least 40 states have introduced bills this year to regulate artificial intelligence, and half a dozen have adopted resolutions or introduced legislation targeting the technology, according to the National Conference of State Legislatures.
Source: AFP
