Pic: Reuters
President Donald Trump ordered all federal agencies to immediately stop using AI tools made by Anthropic, escalating a dispute over the company’s refusal to grant the Pentagon unrestricted access to its systems, March 1, 2026. The move follows the Defense Department’s designation of Anthropic as a “supply chain risk” after the company declined to approve certain classified and military uses of its technology.
Defense Secretary Pete Hegseth formally labeled Anthropic a supply chain risk and directed a six-month phase-out of its tools from government systems. The designation also bars military contractors from engaging in commercial activity with the company during the review period.
The standoff centers on a proposed agreement that would have permitted “any lawful use” of Anthropic’s AI platforms, including potential classified and military applications. According to administration officials, the Pentagon sought broad operational flexibility as it expands artificial intelligence integration across defense systems.
Anthropic rejected uses involving mass domestic surveillance and fully autonomous offensive weapons. In a company statement, Anthropic called the designation “legally unsound” and a “dangerous precedent for any American company that negotiates with the government.”
“No amount of intimidation or punishment … will change our position on mass domestic surveillance or fully autonomous weapons,” the company said, adding that it has not received direct official notice from the White House or Pentagon and intends to challenge the designation in court.
Trump struck a confrontational tone in announcing the directive.
“We don’t need it, we don’t want it, and will not do business with them again,” Trump said. He warned the company to cooperate during the phase-out or face “major civil and criminal consequences,” adding he would use the “Full Power of the Presidency” if necessary.
Hegseth defended the move as a matter of national security.
“Any company supporting the Department of Defense must agree to any lawful use of its tools,” he said, framing Anthropic’s refusal as a compliance issue rather than a policy dispute.
A former Department of Defense official, speaking on condition of anonymity because of the sensitivity of the matter, described the legal basis for invoking the Defense Production Act or applying a supply chain designation as “extremely flimsy.” The former official said Anthropic’s financial position and public stance may limit the Pentagon’s leverage.
The Pentagon contract at issue is valued at roughly $200 million. Anthropic’s valuation is estimated at $380 billion, underscoring the limited financial exposure the company faces from losing government business.
The confrontation also drew reaction from Sam Altman, chief executive officer of OpenAI, a competitor that confirmed it recently reached an agreement with the Defense Department for use of its AI systems on classified cloud networks.
Altman said OpenAI maintains similar “red lines” and would reject military uses that are “unlawful or unsuited to cloud deployments, such as domestic surveillance and autonomous offensive weapons.” He characterized the issue as an industrywide challenge balancing national security demands with ethical safeguards.
The episode marks the first public use of a “supply chain risk” designation against a major U.S. artificial intelligence company and could set a precedent for how far the executive branch may go in compelling private technology firms to support military AI initiatives.
Legal experts said the dispute could test the boundaries between executive authority and corporate autonomy, particularly as artificial intelligence becomes more embedded in defense planning and operations.
Anthropic said it remains willing to work with the government under terms consistent with its policies but signaled it is prepared for a protracted legal fight.
The six-month phase-out is scheduled to begin immediately.