Artificial intelligence company Anthropic is suing to stop the Trump administration from pursuing what it calls an “unlawful retaliatory campaign” after the administration denied unrestricted military use of its technology.
Anthropic on Monday asked a federal court to reverse last week’s Pentagon decision to designate the artificial intelligence company as a “supply chain risk.” The company is also seeking to reverse President Donald Trump’s order directing federal employees to stop using the company’s AI chatbot, Claude.
The legal challenge intensifies an unusual public debate over how AI can be used for war and mass surveillance. The controversy also involves Anthropic’s technology industry rivals, particularly ChatGPT maker OpenAI, which signed its own contract to work with the Department of Defense just hours after the government punished Anthropic for its stance.
Anthropic filed two separate lawsuits Monday in federal court in California and the U.S. Court of Appeals in Washington, D.C., each challenging various aspects of the government’s actions against the San Francisco-based company.
“These actions are unprecedented and illegal,” Anthropic’s lawsuit states. “The Constitution does not allow the government to use its vast powers to punish corporations for protected speech. There is no federal law that authorizes the actions taken here. Antropics looks to the judiciary as a last resort to vindicate its rights and thwart the administration’s unlawful retaliatory efforts.”
The Pentagon declined to comment Monday, citing its policy not to comment on matters pending in litigation.
Anthropic said it seeks to restrict its technology from being used for mass surveillance of Americans or fully autonomous weapons. Defense Secretary Pete Hegseth and other senior officials publicly insisted that the company should accept “all lawful” uses of Claude, threatened punishment if Anthropic did not comply, and criticized the company and its CEO Dario Amodei on social media.
Designating the company as a supply chain risk would cut off Anthropic’s defense efforts, a power designed to prevent foreign adversaries from harming national security systems. This is the first known use of this designation by the federal government for a U.S. company. “It was necessary to protect national security,” Hegseth said in a March 4 letter to Anthropic, according to Anthropic’s complaint.
President Trump also said he would order federal agencies to stop using Claude, but gave the Pentagon six months to phase out the product, which is deeply integrated into sensitive military systems, including those used in the Iran war.
Anthropic’s lawsuit also names other federal agencies, including the Treasury Department and the State Department, after officials ordered employees to stop using Claude.
Anthropic is making some strong First Amendment and due process claims in a lawsuit that has “escalated beyond comprehension,” said Michael Pastor, a professor at New York Law School who previously helped craft technology agreements as New York City’s general counsel.
“I have never seen an incident like this,” the pastor said. “I had no idea that I would be threatening to essentially destroy the company when negotiations were going through a rough patch.”
Even as it counters the Pentagon’s actions, Anthropic has sought to convince companies and other government agencies that the Trump administration’s supply chain risk designation is narrow, affecting only cases where military contractors use Claude for Pentagon work.
For Anthropic, a privately held company, making this distinction is very important. That’s because most of the $14 billion in revenue expected this year will come from businesses and government agencies that use Claude for computer coding and other tasks. More than 500 customers pay Anthropic at least $1 million a year for Claude, according to a recent investment announcement that valued the company at $380 billion.
“While seeking judicial review does not change our long-standing commitment to leveraging AI to protect national security, this is a necessary step to protect our business, customers, and partners,” Anthropic said in a statement Monday.
The lawsuit positions AI safety and “positive outcomes for humanity” as important to Anthropic’s mission since its founding in 2021 by Amodei and six other former OpenAI employees.
The company said in its lawsuit that its use policy has always prohibited “lethal autonomous warfare without human oversight or surveillance of the American people as a whole.” Anthropic said it has not tested Claude in these applications and has no confidence that its product “will function reliably or safely when used to support lethal autonomous warfare.”
At the same time, the military has been able to use Claude in ways that civilians cannot, such as analyzing military operations and “lawfully collected foreign intelligence information.”
Until recently, Anthropic was the only company among its tech industry peers approved to supply AI models to classified military systems. The controversy has led the Pentagon to consider moving Claude’s work to Google’s Gemini, OpenAI’s ChatGPT and Elon Musk’s Grok.
Anthropic’s lawsuit alleges that the Trump administration’s actions seek to damage the company’s reputation, “put hundreds of millions of dollars at risk” in contracts with other companies, and “destroy the economic value created by one of the world’s fastest-growing private companies.”
On the contrary, the battle has boosted Anthropic’s reputation among some customers and technology insiders who support the company in not bowing to pressure from the Trump administration. Amodei’s moral stance became even more pronounced when his bitter rival, OpenAI CEO Sam Altman, tried to replace Claude at the Pentagon with ChatGPT, a move that Altman later acknowledged was hasty and opportunistic.
Consumer downloads of Claude have skyrocketed, making it more popular than the better-known ChatGPT and Gemini for the first time.
How companies set guardrails continues to impact the race to retain talent in the AI industry. Caitlin Kalinowski, OpenAI’s head of robotics, has resigned over OpenAI’s contract with the Department of Defense.
“This was not an easy decision,” Kalinowski wrote on social media over the weekend. “AI plays an important role in national security, but surveillance of Americans without judicial oversight or lethal autonomy without human permission deserves full consideration.”
OpenAI and a separate group of more than 30 leading AI developers at Google, including Jeff Dean, Google’s chief scientist and head of AI research, filed a legal brief in support of Anthropic on Monday.
“National security is not protected by recklessly designating the military’s U.S. technology partners as ‘supply chain risks’ or by suppressing public debate about the safety of AI,” the workers’ filing, which said they were acting in their personal capacity, said.
Copyright © 2026 Associated Press. Unauthorized reproduction is prohibited. This website is not directed to users within the European Economic Area.
