California Gov. Gavin Newsom on Sept. 29 signed the nation’s first law requiring major AI developers to disclose their safety protocols, a move that could set the stage for national standards, Politico reported.
Here are six things to know:
The law, Senate Bill 53, mandates that some AI companies publicly reveal their security plans, report major safety incidents to the state and protect whistleblowers. It also establishes the groundwork for a state-run cloud computing system called CalCompute, according to the Sept. 29 Politico report.
The law places California ahead of efforts in other states, including New York, where similar legislation awaits Gov. Kathy Hochul’s approval.
Gov. Newsom signed the bill days after suggesting he would approve it while speaking at a United Nations event alongside former President Bill Clinton.
The measure’s passage follows a veto last year of a broader proposal from state Sen. Scott Wiener, a San Francisco Democrat. Mr. Wiener revised the legislation and gained support from some AI companies, including Anthropic, while others such as OpenAI and Meta did not publicly oppose it, Politico reported.
Tech companies and industry groups expressed mixed reactions. Anthropic co-founder Jack Clark said the company was proud to work with Mr. Wiener to create “practical safeguards,” while OpenAI spokesperson Jamie Radice said the law “created a critical path toward harmonization with the federal government.” Meta spokesperson Christopher Sgro called the measure “a positive step.”
Others criticized the law. The Chamber of Progress, a tech lobbying group, warned it could hurt innovation, while Andreessen Horowitz government affairs head Collin McCune said it risked favoring large players over startups.
The post California enacts 1st US law requiring AI safety disclosures appeared first on Becker’s Hospital Review | Healthcare News & Analysis.
Health IT
