Governor Gavin Newsom of California issued an executive order Monday establishing new requirements for companies that want to do business with the state, mandating they implement protections against artificial intelligence misuse that could generate illegal material, create harmful bias, or violate civil rights.
Under the new directive, state agencies must add watermarks to any images or videos created using artificial intelligence technology, following state-issued guidelines designed to combat misinformation. The order also addresses federal security concerns by allowing California to conduct independent evaluations of companies flagged as supply chain risks by the federal government, potentially keeping them as contractors if the state determines they pose no threat.
The timing of this order follows the Pentagon’s recent decision to designate artificial intelligence company Anthropic as a formal supply-chain security risk, preventing government contractors from utilizing the firm’s technology in military-related work.
California’s Department of General Services and Department of Technology have been given 120 days to develop recommendations for new vendor certification requirements related to AI technology. These certifications would enable companies to demonstrate responsible artificial intelligence governance and public safety protections.
The executive order reflects California’s commitment to maintaining an independent regulatory approach, despite pressure from some Republican legislators to align with federal oversight and regulations.
Earlier this year in February, California Attorney General Rob Bonta explained in a Reuters interview that his office is building internal capabilities through its “AI oversight, accountability and regulation program.”
