Anthropic has sued to block the Pentagon from placing it on a U.S. national security blacklist, escalating the artificial intelligence lab’s high-stakes battle with U.S. President Donald Trump’s administration over restrictions on the use of its technology.
Anthropic said in its lawsuit Monday that the designation was illegal and violated its free speech and due process rights. The filing in federal court in the U.S. state of California asked a judge to undo the designation and block federal agencies from enforcing it.
Recommended stories
list of 4 itemsend of list
“These actions are unprecedented and illegal. The Constitution does not allow the government to exercise its enormous power to punish a company for its protected speech,” Anthropic said.
On Thursday, the Pentagon imposed a formal supply chain risk designation on Anthropic, limiting its use of a technology that Reuters news agency reported, citing an anonymous source, was being used for military operations in Iran.
US Defense Secretary Pete Hegseth appointed Anthropic after the startup refused to remove barriers against using its AI for autonomous weapons or domestic surveillance. The two sides had been in increasingly contentious talks over those limitations for months.
Trump and Hegseth said there would be a six-month phase-out.
The company is also seeking to undo Trump’s order directing federal employees to stop using its AI chatbot, Claude.
The legal challenge intensifies an unusually public dispute over how AI can be used in warfare and mass surveillance, a dispute that has also dragged in Anthropic’s tech industry rivals, notably OpenAI, which reached its own agreement to work with the Pentagon just hours after the government punished Anthropic for its stance.
Anthropic filed two separate lawsuits on Monday, one in federal court in California and another in the federal appeals court in Washington, D.C., each challenging different aspects of the government’s actions against the company.
Anthropic officials said the lawsuit does not preclude reopening negotiations with the U.S. government and reaching a deal. The company has said it does not want to fight the US government. The Pentagon said it would not comment on the litigation. Last week, a Pentagon official said the two sides were no longer in active talks.
Threat to business
The designation poses a major threat to Anthropic’s business with the government, and the outcome could determine how other AI companies negotiate restrictions on military use of their technology, although the company’s CEO Dario Amodei clarified Thursday that the designation was “limited in scope” and companies could still use its tools on projects unrelated to the Pentagon.
Trump and Hegseth’s actions on Feb. 27 came after months of talks with Anthropic about whether the company’s policies could limit military action and shortly after Amodei met with Hegseth in hopes of reaching a deal.
Anthropic said it was seeking to restrict the use of its technology to two high-level uses: mass surveillance of Americans and fully autonomous weapons. Hegseth and other officials publicly insisted that the company must accept “all legal uses” of Claude and threatened to punish them if Anthropic did not comply.
Designating the company as a supply chain risk disrupts Anthropic’s defense work using authority that was designed to prevent foreign adversaries from damaging national security systems. It was the first time the federal government was known to have used the designation against an American company.
The Pentagon said American law, not a private company, would determine how to defend the country, and insisted on having full flexibility in using AI for “any lawful use,” saying Anthropic’s restrictions could endanger American lives.
Anthropic said that even the best AI models were not reliable enough for fully autonomous weapons and that using them for that purpose would be dangerous.
After Hegseth’s announcement, Anthropic said in a statement that the designation would be legally flawed and set a dangerous precedent for companies dealing with the government. The company said it would not be influenced by “intimidation or punishment.”
Last week, Amodei also apologized for an internal memo published Wednesday by tech news site The Information. In the memo, released Feb. 27, Amodei said Pentagon officials did not like the company in part because “we have not praised Trump in a dictatorial style.”
Even as it fights the Pentagon’s actions, Anthropic has tried to convince companies and other government agencies that the Trump administration’s penalty is limited and only affects military contractors when they use Claude to work for the Department of Defense.
Making that distinction clear is crucial for privately held Anthropic because most of its projected $14 billion in revenue this year comes from companies and government agencies that use Claude for computer coding and other tasks. More than 500 clients are paying Anthropic at least $1 million a year for Claude, according to a recent investment announcement that valued the company at $380 billion.
