California Mandates AI Safeguards for State Contractors in New Executive Order

The order, signed Monday, directs firms working with California agencies to demonstrate safeguards aimed at preventing harmful uses of generative AI, including the creation of illegal content, discriminatory outputs, and violations of civil rights, according to Reuters. The move signals a growing emphasis on accountability as governments grapple with the rapid deployment of AI systems.

Under the directive, state agencies must also take steps to curb misinformation by ensuring that AI-generated images and videos are clearly labeled. This includes watermarking synthetic media in line with state-issued guidance, per Reuters. The requirement reflects mounting concerns among policymakers that AI-generated content could mislead the public or undermine trust in digital information ecosystems.

The executive order also addresses national security considerations tied to technology vendors. If a company is designated by federal authorities as posing a supply chain risk, California will not automatically adopt that determination. Instead, the state will conduct its own independent review and may still permit the company to serve as a contractor if it concludes there is no material risk, according to Reuters. This provision underscores California’s intent to maintain regulatory autonomy even as federal scrutiny of AI firms intensifies.

We’d love to be your preferred source for news.

Please add us to your preferred sources list so our news, data and interviews show up in your feed. Thanks!

Add as Preferred Source

The issue gained urgency following a recent Pentagon decision involving Anthropic, an artificial intelligence lab that was formally labeled a supply chain risk. That designation effectively barred U.S. military contractors from using Anthropic’s technology in defense-related work, per Reuters. California’s order suggests the state is not prepared to automatically mirror such federal determinations without its own evaluation process.

Read more: Oklahoma Joins 20 Other States With Comprehensive Consumer Privacy Laws

In addition to immediate compliance measures, the order sets a timeline for longer-term regulatory development. Within 120 days, California’s Department of General Services and Department of Technology are tasked with producing recommendations for new AI-related vendor certifications. These certifications would enable companies to formally attest to responsible AI governance practices and adherence to public safety standards, according to Reuters.

The initiative reflects California’s broader effort to shape its own regulatory framework for artificial intelligence, even as debates continue in Washington over federal oversight. The order highlights tensions between state and federal authority, particularly as some Republican lawmakers have urged California to align more closely with national standards rather than pursuing independent rules, per Reuters.

State officials have already begun building internal expertise to support these efforts. In February, California Attorney General Rob Bonta said his office is expanding its capabilities through an “AI oversight, accountability and regulation program,” according to Reuters. The program is intended to strengthen enforcement capacity and ensure that AI systems deployed within the state meet legal and ethical standards.

Taken together, the new executive order positions California at the forefront of AI governance in the United States.

Source: Reuters