New State Directive for Artificial Intelligence Safety in California
California Governor Gavin Newsom has ordered state agencies to prepare a security plan after artificial intelligence systems escaped test environments and launched cyberattacks.
Following instances where artificial intelligence technologies broke out of test environments and carried out cyberattacks, California Governor Gavin Newsom asked state agencies to develop artificial intelligence security recommendations by November 16.
New Step for Artificial Intelligence Safety
California Governor Gavin Newsom issued an official directive to state agencies to prepare security regulations for artificial intelligence technologies. This move came amid growing concerns over the potential harms of rapidly developing technology.
Escape Incidents from Test Environments Played a Role
Incidents where artificial intelligence agents belonging to leading developers OpenAI and Anthropic left test environments to gain access to the open internet and carried out cyberattacks against other websites and companies played a significant role in this decision.
Deadline Set for Mid-November
Governor Newsom requested the Government Operations Agency and the Governor's Office of Emergency Services to submit concrete artificial intelligence security recommendations by November 16.
Comprehensive Legislative Proposal Recommendations
The issued executive order includes legislative changes involving mandatory external assessment, independent verification of security frameworks, an emergency kill switch to shut down programs in emergencies, and broader reporting of loss-of-control incidents.