California’s AI bill gains industry support

Around 120 current and former employees from AI giants like OpenAI, Anthropic, DeepMind, and Meta have publicly voiced their support for California’s new AI regulation bill, SB 1047. The bill, which includes whistle-blower protections for employees revealing the risks in AI models, aims to impose stronger regulations on developing powerful AI technologies. Supporters argue that these measures are crucial to prevent potential threats such as cyberattacks and the misuse of biological weapons.

California’s SB 1047 has already passed the State Assembly and Senate and is awaiting Governor Gavin Newsom’s decision, with a deadline set for 30 September. Notably, high-profile signatories of the letter backing the bill include Geoffrey Hinton, a Turing Award winner, and Jan Leike, a former OpenAI alignment lead, signalling wide support from influential figures in the tech world.

Proponents of the bill believe AI companies should be responsible for testing and ensuring their models don’t pose significant harm. They argue that regulations are essential to safeguard critical infrastructure and prevent AI misuse. Despite its limitations, experts like Harvard’s Lawrence Lessig have called the bill a ‘solid step forward’ in managing AI risks.

However, not everyone agrees. OpenAI and other major tech organisations, including the US Chamber of Commerce and the Software and Information Industry Association, oppose the bill, claiming it would stifle innovation in the fast-moving AI sector. Tech industry advocates argue that over-regulation may hinder the development of cutting-edge technologies.