Artificial intelligence firm Anthropic has filed first-of-its-kind lawsuits against the U.S. government in response to the Department of War designating the company behind Claude a “supply-chain risk” – a decision that effectively blacklists it from all federal and defense-related projects.
Anthropic has been used by the U.S. government and military since 2024, with Claude becoming the first advanced AI model deployed in government agencies for classified work.
Anthropic Sues Trump Administration Over Supply-Chain Risk Label
The company filed complaints against the War Department (formerly Department of Defense) on Monday in California and Washington, D.C., after Amodei and Defense Secretary Pete Hegseth publicly sparred last week over Anthropic’s refusal to allow the military unrestricted access to its AI systems.
Anthropic has sued President Donald Trump’s executive office, Secretary Hegseth, Secretary of State Marco Rubio, Secretary of Commerce Howard Lutnick, and 16 federal agencies, including the Department of War, Department of Homeland Security, and the Department of Energy.
The dispute centers on two specific safety guardrails that Anthropic refuses to waive: using its AI for mass domestic surveillance of American citizens and the deployment of lethal autonomous weapons that can act with no humans making targeting and firing decisions.
Negotiations between Anthropic, the first frontier AI firm to deploy models for national security uses, and the Pentagon fractured in late February. Hegseth reportedly demanded that the company allow “all lawful uses” of its technology, arguing that private corporations should not hold veto power over military operations. However, when CEO Dario Amodei refused the request, the Trump administration reiterated by designating the company a “supply-chain risk” – a label reserved for foreign adversaries and currently applied to Chinese tech giants Huawei and ZTE.
The designation also requires any company or agency that works with the Pentagon to certify that they do not use Anthropic’s products. Claude is one of the most popular large-language models (LLMs), with its ‘Claude Code’ being an almost ubiquitous part of work done by some of the biggest tech firms in the US, including Google, Meta, Amazon, and Microsoft.
While those companies also work with the government, last week, Microsoft, Amazon, and Google said they would continue to use Claude outside of any work for the War Department. However, Anthropic is set to lose much of its government business.
The impact of the supply-chain risk label was immediate. Shortly after the Pentagon’s announcement on March 4, President Trump issued a directive ordering all federal agencies to “immediately cease” all use of Anthropic’s technology. The White House granted a six-month phase-out period for existing military contracts with the firm.
Anthropic Challenges Trump’s Federal Agency Phase-Out Order
Anthropic claims that it has been “irreparably” harmed by Trump and Hegseth’s comments, with the government’s move jeopardizing “hundreds of millions of dollars” in current and future contracts with private parties. The company, which was recently valued at $380 billion, argues that the supply-chain risk tag hindered its reputation and threatens its ability to compete in the global market.
“The Constitution does not allow the government to wield its enormous power to punish a company for its protected speech,” Anthropic said in the lawsuit. “No federal statute authorizes the actions taken here.”
In a rare show of industry solidarity, nearly 40 Google and OpenAI employees filed an amicus brief with the court supporting Anthropic and its efforts to limit improper uses of AI, offering their expertise on the dangers posed by the technology being used at scale. The signatories argued that while they are fierce competitors, the government’s use of national security labels to settle contractual disputes over AI safety sets a dangerous precedent for the entire tech sector.
Meanwhile, the Pentagon wasted no time replacing Claude. Within hours of blacklisting Anthropic, OpenAI reportedly signed a multi-million dollar deal to provide its ChatGPT models for classified networks, national labs, and national security uses like intelligence analysis and cyber operations, while agreeing to the military’s broader usage terms.
Anthropic is not seeking monetary damages from its lawsuit, but is asking the court to immediately declare that Trump’s directive “exceeds the president’s authority” and violates the Constitution, and reject the company’s label as a supply-chain risk. However, the White House remains defiant, with spokesperson Liz Huston telling the BBC that Anthropic is a “radical left, woke company” attempting to control military activity.
“Under the Trump Administration, our military will obey the United States Constitution – not any woke AI company’s terms of service,” Houston said.
