OpenAI announces agreement with Department of Defense after President Trump bans Anthropic

Updated on February 28, 2026 at 2:46 AM CST

President Trump ordered the U.S. government to halt use of artificial intelligence company Anthropic, and the Pentagon moved Friday to designate the company a national security risk, as the high-stakes battle over the military’s use of AI rapidly escalates.

Hours after the president’s announcement, rival company OpenAI announced it had signed a deal with the Department of Defense to provide its AI technology to sensitive networks.

The administration’s decision ends a bitter dispute between Anthropic and the Department of Defense over whether Anthropic can be barred from using its tools for mass surveillance of U.S. citizens or to power autonomous weapons systems as part of a military contract worth up to $200 million.

“The left-wing lunatics at Anthropic made a disastrous mistake in trying to superpower the Department of the Army and force it to follow the Code of Service instead of the Constitution,” Trump wrote in a post on Truth Social. “Therefore, I am directing all federal agencies of the United States Government to immediately stop using Anthropic’s technology. We don’t need it. We don’t want it. We will never do business with them again!”

He said Anthropic’s products would be phased out over a six-month period.

President Trump’s announcement came about an hour before the deadline set by the Pentagon for Antropic to withdraw. Shortly after the deadline passed, Secretary of Defense Pete Hegseth said he was classifying Anthropic as a supply chain risk to national security and blacklisting it from working with the U.S. military or contractors.

“In conjunction with the President’s direction to cease all use of Anthropic’s technology for the federal government, I am directing the Department of the Army to designate Anthropic as a supply chain risk to national security. Effective immediately, any contractor, supplier, or partner doing business with the U.S. military may not engage in any commercial activity with Anthropic,” Hegseth posted on X, using the Department of Defense’s rebranding as “Department of the Army.” “Anthropic will continue to provide its services to the Department of the Army for a period of no more than six months to allow for a seamless transition to a better, more patriotic service.”

Anthropic said it would challenge the supply chain risk designation in court.

“We believe this designation is legally unsound and sets a dangerous precedent for U.S. companies negotiating with the government,” the company said in a statement Friday night.

Anthropic also took issue with Mr. Hegseth’s comments that anyone doing business with the U.S. military must cut off all business with Anthropic. “The Secretary has no legal authority to support this statement,” the company said. Under federal law, Anthropic’s designation as a supply chain risk “applies only to the use of Claude as part of a contract with the Department of the Army. It does not affect how the contractor uses Claude to provide services to other customers.”

The company said it made a “good faith effort” to reach an agreement with the Pentagon over months of negotiations and “made clear that we support all lawful uses of AI for national security, with two narrow exceptions at issue.” “To our knowledge, these exceptions have not affected a single government mission so far,” Antropic said.

The group said there were two reasons for its opposition to their use. “First, we do not believe that today’s Frontier AI models are reliable enough to be used in fully autonomous weapons. Allowing current models to be used in this manner will put American warfighters and civilians at risk. Second, we believe that large-scale domestic surveillance of Americans constitutes a violation of their fundamental rights.”

In a post on X announcing competitor OpenAI’s deal with the Department of Defense, company CEO Sam Altman previously cited similar concerns and said the agreement with the government includes safeguards like those sought by Anthropic.

“Two of our most important security principles are the prohibition of domestic mass surveillance and human responsibility for the use of force, including autonomous weapons systems,” he said. “The DoW agrees with these principles and has reflected them in our laws and policies, and we have included them in our agreement.”

Ban issued as Anthropic plans IPO

Pentagon officials had given Anthropic a deadline of 5:01 p.m. ET on Friday to lift domestic restrictions on its AI model, Claude, for use in mass surveillance and fully autonomous weapons, or lose the contract. The Pentagon has said it does not intend to use AI in that way, but has asked AI companies to allow their models to be used “for all lawful purposes.”

The government also threatened to invoke the Korean War-era Defense Production Act to force Anthropic to permit use of its tools, while also warning that it would designate Anthropic as a supply chain risk.

In a post carrying out the latter threat, Hegseth said Anthropic “not only provided a master class in arrogance and betrayal, but also a textbook example of how not to do business with the U.S. government and Department of Defense.” He accused the company of trying to “grasp a veto over U.S. military operational decisions.”

He said the Department of the Army remains unwavering in its position that it “must have full and unrestricted access to Anthropic’s models for all lawful purposes in the defense of the Republic.”

“America’s warfighters will never be held hostage to the ideological whims of Big Tech companies. This decision is final,” Hegseth concluded.

The government’s ban comes amid increased scrutiny of Anthropic, which has a market capitalization of $380 billion and plans to go public this year.

The $200 million Pentagon contract is a relatively small portion of Anthropic’s $14 billion in revenue, but it’s unclear how the friction with the administration will affect investors or other contracts the company needs to license its AI models to non-government partners.

Anthropic CEO Dario Amodei said the company’s valuation and revenue have only increased since the company clashed with the Trump administration over how to deploy AI on the battlefield.

Whether AI companies can set limits on how the government uses their technology has emerged as a major point of contention between Anthropic and the Trump administration in recent months.

Amodei said Thursday that the company would not be intimidated by the Pentagon’s threats. “We cannot in good conscience comply with their demands,” he said in a lengthy statement.

“Antropic understands that military decisions are made by the Department of the Army, not private companies. We have never objected to specific military operations, nor have we sought to limit the use of technology in military operations.” For this purpose But he added that domestic mass surveillance and fully autonomous weapons are applications that are “way beyond what can be safely and reliably implemented with today’s technology.”

Emile Michael, the Pentagon’s assistant secretary for research and engineering, accused Amodei of lying and having a “god complex” in a Thursday post on X.

“He wants nothing more than to seek personal control of the U.S. military and is comfortable putting our nation’s security at risk,” Michael wrote. “@DeptofWar will always abide by the law, but we will not bow to the whims of commercial technology companies,” he wrote.

In an interview with CBS News late Thursday, Michael said federal law and Pentagon policy already prohibit the use of AI in domestic mass surveillance or autonomous weapons.

“At some level, you have to trust the military to do the right thing,” he said.

OpenAI expressed similar concerns

OpenAI CEO Altman said earlier Friday that he sympathizes with Anthropic’s “red lines” that limit the military use of AI.

OpenAI, Google, and Elon Musk’s xAI also have contracts with the Department of Defense, agreeing to allow their AI tools to be used in any “legal” scenario. Earlier this week, xAI became the second company after Anthropic to be approved for use in classified environments.

Altman told CNBC on Friday morning that it is important for companies to cooperate with the military “as long as they comply with legal protections,” and that there are “some red lines” that “we share with Anthropic and that other companies have agreed to on their own” that should not be crossed.

In an internal memo sent to staff Thursday night, Altman said he is seeking to negotiate an agreement with the Pentagon to deploy the model on classified systems with exemptions that would prevent OpenAI from being used for surveillance purposes in the United States or powering autonomous weapons without human approval, according to a person familiar with the message who was not authorized to speak publicly. of wall street journal Altman’s memo was first reported to staff.

The Pentagon did not respond to a request for comment on Altman’s remarks.

Independent experts say the conflict is highly unusual in the world of Pentagon contracting.

“This is clearly different,” said Jerry McGinn, director of the Industrial Infrastructure Center at the Center for Strategic and International Studies think tank in Washington, D.C. He noted that Pentagon contractors typically cannot tell the agency how their products and services will be used. “Otherwise, you’ll be negotiating use cases for each contract, and that’s not a reasonable expectation.”

At the same time, McGinn pointed out that artificial intelligence is a new and largely untested technology. “This is a very unusual and very public battle,” he said. “I think this reflects the nature of AI.”

NPR’s Bobby Allyn contributed to this report.

Copyright 2026 NPR