Artificial intelligence firm Anthropic has filed a lawsuit seeking to block the U.S. Pentagon from placing it on a national security blacklist, escalating a major dispute between the company and the U.S. military over restrictions on how its technology can be used. According to Reuters, the lawsuit was filed Monday in federal court in California and asks a judge to overturn the designation and prevent federal agencies from enforcing it.
Anthropic argues that the Pentagon’s decision to impose the designation is unlawful and violates the company’s constitutional rights, including free speech and due process protections. In its legal filing, the company said the government cannot use its authority to penalize a firm for maintaining policies that limit how its artificial intelligence systems are deployed.
The conflict intensified last week when the Pentagon applied a formal supply-chain risk designation to Anthropic, effectively restricting the use of its technology within certain government operations. Reuters previously reported that Anthropic’s tools had been used in military-related activities connected to operations in Iran, and that tensions had been growing for months between the company and the Defense Department over safeguards embedded in its systems.
The designation was authorized by U.S. Defense Secretary Pete Hegseth after Anthropic declined to remove guardrails designed to prevent its AI models from being used for autonomous weapons or for domestic surveillance. The company has argued that even advanced AI systems are not reliable enough to safely operate fully autonomous weapons and warned that such use could pose significant risks.
Pentagon officials have taken a different view, saying that U.S. law, rather than corporate policies, should determine how the military defends the country. They have insisted on retaining full flexibility to deploy artificial intelligence tools for what they describe as “any lawful use,” arguing that limitations imposed by private companies could endanger American lives.
Anthropic said in its lawsuit that the government’s actions were “unprecedented and unlawful,” adding that the Constitution does not permit authorities to punish a company for its protected speech or policy decisions. Despite the legal challenge, company officials indicated that the lawsuit does not rule out the possibility of renewed negotiations with the U.S. government or a potential settlement.
The dispute could have significant consequences for Anthropic’s business with federal agencies and may influence how other AI developers negotiate restrictions on military applications of their technology. According to Reuters, Anthropic’s chief executive Dario Amodei has said the Pentagon’s designation currently has a relatively narrow scope, meaning companies can still use the firm’s tools in projects unrelated to the Defense Department.
The conflict has also drawn in the White House. Reuters reported that President Donald Trump has directed federal agencies to stop working with Anthropic, with officials announcing a six-month phase-out period for government reliance on the company’s technology. The move has prompted concern among Anthropic’s investors, which include major technology companies such as Google and Amazon, as they attempt to limit the fallout from the dispute.
The tensions follow months of discussions between Anthropic and Pentagon officials about whether the company’s policies could restrict military operations. Shortly before the designation was announced on February 27, Amodei met with Hegseth in an attempt to reach an agreement, but the talks ultimately collapsed.
Adding to the controversy, Amodei recently apologized for comments in an internal memo that were published by the technology news outlet The Information. In the memo, he suggested that some Pentagon officials disliked the company partly because it had not offered what he described as “dictator-style praise” for Trump.
The legal fight unfolds as the Defense Department expands its partnerships with leading artificial intelligence developers. Over the past year, Reuters reported, the Pentagon has signed agreements worth up to $200 million each with several major AI laboratories, including Anthropic, OpenAI, and Google. Shortly after Anthropic was blacklisted, OpenAI announced a separate deal to integrate its technology into the Defense Department’s network, with its chief executive Sam Altman emphasizing that the partnership would maintain human oversight over weapon systems and oppose mass surveillance of Americans.
