Governor Gavin Newsom signed an executive order Monday that puts California on a collision course with the Trump administration over who gets to decide which artificial intelligence companies are trustworthy — and made clear the state intends to make that determination on its own terms.
The order, signed March 30, arrives at a moment when federal AI policy and California’s long-established position as the center of the domestic AI industry are pulling hard in opposite directions. It is as much a governing document as it is a declaration of California’s intent to hold its ground.
The Dispute That Triggered the Order
In July 2025, Anthropic and the Pentagon entered into a contract under which Claude became the first frontier model approved for use on classified networks. As part of the contract, the Pentagon agreed to abide by Anthropic’s acceptable use policy. The Pentagon later sought to renegotiate those terms, insisting that Anthropic allow the military to use Claude “for all lawful purposes” without limitation. The parties engaged in weeks of failed negotiations.
Anthropic’s CEO made clear the company would not drop two demands: no use of its artificial intelligence for fully autonomous weapons — meaning AI, not humans, making final battlefield targeting decisions — and no mass domestic surveillance. When Anthropic did not agree to the Pentagon’s deadline, Defense Secretary Pete Hegseth posted a statement declaring the company a supply chain risk and telling every defense contractor they could not use the company’s AI. “America’s warfighters will never be held hostage by the ideological whims of Big Tech,” Hegseth said.
The label had previously been used only for companies seen as connected to foreign adversaries. Some military contractors were already cutting ties with Anthropic. Lockheed Martin said it would “follow the President’s and the Department of War’s direction” and look to other providers of large language models. Hours after the Pentagon punished Anthropic, OpenAI announced a deal to effectively replace Anthropic with ChatGPT in classified military environments.
A federal judge in California indefinitely blocked the Pentagon’s effort, ruling that the measures “do not appear to be directed at the government’s stated national security interests.” U.S. District Judge Rita Lin wrote: “Nothing in the governing statute supports the Orwellian notion that an American company may be branded a potential adversary and saboteur of the U.S. for expressing disagreement with the government.” She found the designation likely violated Anthropic’s First Amendment and due process rights.
What Newsom’s Order Does
The next time the federal government labels a business a supply-chain risk, as the Department of Defense did to San Francisco-based Anthropic, the state of California will review that designation and make its own decision about whether to do business with them. The order instructs California’s Chief Information Security Officer to review any federal designations of AI companies as supply chain risks, and gives the state the authority to overrule those designations if it deems them improper.
The order also tasks the Government Operations Agency, working with the California Department of Technology and the Department of General Services, with recommending reforms to contractor responsibility provisions. If the Chief Information Security Officer concludes a federal designation is improper, state guidance will be issued ensuring that departments and agencies can continue to easily procure from that company.
The order directs the Government Operations Agency to develop a plan for new state contracting processes and best practices that vet companies based in part on how they attest to and explain their policies and safeguards to protect the public. Areas of interest include exploitation and distribution of illegal content, use of models that promote harmful bias, and protection of civil rights.
The order also directs the California Department of Technology to create recommendations and best practices for watermarking AI-generated images or manipulated video consistent with state law — the first of its kind nationwide.
California’s Position in the National AI Landscape
The Newsom order does not arrive in a vacuum. It builds on California’s years of AI-related legislation and reflects the state’s unique position in the national technology ecosystem.
As the world’s fourth-largest economy and home to 33 of the top 50 privately held AI companies globally, California is one of the few states with the size and power to go head-to-head with the federal government on this question. That portfolio includes not just Anthropic but OpenAI, Google DeepMind, Meta AI, and dozens of companies building the infrastructure on which AI-dependent industries now run.
Since the current legislative session began in December 2024, more than 50 bills that regulate AI are being considered in the California legislature according to CalMatters’ Digital Democracy database. Federal legislative gridlock and the Trump administration’s hostile stance toward AI regulation mean that California, where Democrats hold a supermajority, is one of the few places in the country where meaningful AI regulation may actually happen.
The Federal Preemption Question
The order puts California directly at odds with the White House’s national AI policy framework, released earlier this month.
The Trump document, which lays out legislative recommendations to Congress, calls for preempting state AI laws that the administration considers “burdensome.” It argues that AI development is “an inherently interstate phenomenon with key foreign policy and national security implications,” and that states should not be permitted to regulate it. Newsom’s order fires back directly.
Newsom said the order comes in response to moves by the Trump administration to limit state regulation of AI in favor of a single nationwide approach, following lobbying by large tech companies. “While others in Washington are designing policy and creating contracts in the shadow of misuse, we’re focused on doing this the right way,” Newsom stated.
Whether Congress acts on the White House’s recommendations and passes legislation to override state AI laws remains the defining legal question. If it does, California’s order — and the broader body of AI regulation the state has built — could face direct preemption challenges. If it does not, California’s framework could become the de facto regulatory standard for the industry.
The Broader Stakes for AI Companies
Within 120 days, California’s Department of General Services and Department of Technology are expected to propose a certification framework requiring vendors to attest to responsible AI governance and public safety protections. This applies to all companies seeking state contracts — a significant market given California’s scale as a buyer of technology services.
To offer a financial incentive to AI companies alongside the new restrictions, Newsom’s order directs the state to build a public-facing AI tool that helps Californians navigate government services organized around life events like job searching, starting a business, or applying for disaster relief. This could lead to significant licensing revenue for AI companies supplying the underlying technology.
For Silicon Valley companies navigating a moment when the federal government is rewriting the rules of AI procurement in real time, California’s move offers both an alternative framework and a signal. Sacramento is not going to follow Washington’s lead on AI without its own review — and the state’s size gives that position practical consequence.
Newsom said: “California’s always been the birthplace of innovation. But we also understand the flip side: in the wrong hands, innovation can be misused in ways that put people at risk. California leads in AI, and we’re going to use every tool we have to ensure companies protect people’s rights, not exploit them or put them in harm’s way.”
The executive order is the second signed by Newsom specifically to address artificial intelligence. It will not be the last. The 2026 legislative session, the open governor’s race, and the ongoing federal-state standoff over AI governance make California’s regulatory posture toward the technology one of the defining policy stories of the year.
