California-based AI firms are protesting the new AI bill that imposes tighter controls on the sector and potentially stifles innovation. The proposed bill compels firms to adhere to a strict safety framework. This will include creating a âkill switchâ to turn off their powerful AI models as the battle to control the sector escalates. Â
 Also read: Tackling the Threat of Deceptive AI: Researchers Issues Urgent Warming
Some measures proposed under the bill might negatively impact tech firms like Meta, while it will also create restrictions for AI startup firms like OpenAI, Anthropic, and Cohere.
State senate passed the bill last month and now set for a vote from its general assembly in August this year.
The California bill threatens open-source AI models
According to arstechnica, big tech firms must guarantee they will never create models that are harmful to society, like nuclear weapons or aid cyber security attacks.
However, the proposed bill is receiving a backlash from some stakeholders in the sector, arguing this has a knock-on effect on startups who might be forced to leave the state. It may also prevent big firms like Meta from open-sourcing their models.
âIf someone wanted to come up with regulations to stifle innovation, one could hardly do better,â said Andrew NG, a computer scientist who led AI projects at Google and Chinaâs Baidu, and sits on Amazonâs board.
âIt creates massive liabilities for science-fiction risks, and so stokes a fear in anyone daring to innovate.â
Arun Rao, a lead product manager for generative AI at Meta said the âunworkableâ bill would âend open-source (in California).â
The protests are expected, according to Senator Wiener
Democratic state Senator Scott Wiener, who introduced the bill, shrugged off the criticism. Although facing a backlash for being overly restrictive, the legislator indicated the protests were nothing new, coming from the tech sector.
âThis is the tech sector, it doesnât like to have any regulation so its not surprising to me at all that there would be push back.â
Senator Scott Wiener.
âFundamentally I want AI to succeed and innovation to continue but letâs try and get out ahead of any safety risks,â he said.
San Francisco-based non-profit organization Center for AI Safety (CAIS), headed by Dan Hendrycks co-sponsored the California bill.
Also read: âAI Godfatherâ Wants Universal Basic Income for AI Job Losses
Hendrycks supported the bill, saying it could help address one of the problems in the sectors â cutting corners at a time when people also want âsome basic oversight.â
âThere are these competitive pressures that are affecting these AI organizations that incentivize them to cut corners on safety.â
The AI sector has experienced rapid growth, which has also raised concerns about the safety of the technology. Last year, tech billionaire Elon Musk called AI a âthreat to humanity.â
Cryptopolitan reporting by Enacy Mapakame