TLDR

  • Anthropic initiated two legal actions against the Pentagon and other federal agencies on March 9, 2026
  • The Pentagon deemed Anthropic a “supply-chain risk” to national security after the firm refused to remove AI safety safeguards
  • President Trump issued a directive via social media for the entire federal government to cease using Claude, Anthropic’s AI product
  • Anthropic asserts the government’s actions are illegal and infringe on its First Amendment rights and due process
  • OpenAI revealed a new Pentagon agreement shortly following Anthropic’s blacklisting

Anthropic has filed lawsuits against the U.S. Defense Department and multiple federal agencies after the Pentagon placed the AI company on a national security blacklist, as announced on Monday.

The company launched two distinct lawsuits—one in the Northern District of California and another in the D.C. Circuit Court of Appeals. Both challenges target the government’s decision to label Anthropic a “supply-chain risk.”

The conflict stems from disagreements over the military’s use of Anthropic’s AI system, Claude. The Pentagon sought to utilize Claude for “any lawful purpose,” but Anthropic declined to remove guardrails that restricted the AI’s application in autonomous weapons or domestic surveillance.

Defense Secretary Pete Hegseth formally designated Anthropic as a supply-chain risk on February 27, with the company receiving official notification on March 3.

President Trump expanded the measure by posting on social media, ordering the entire federal government to stop using Claude, exceeding the Pentagon’s initial designation.

Anthropic stated the government’s actions are “unprecedented and unlawful,” adding that its “reputation and core First Amendment freedoms are under attack.” The company argues the moves constitute retaliation for protected speech, not a genuine national security decision.

“The Constitution does not permit the government to wield its immense power to penalize a company for its protected speech,” Anthropic noted in its filing.

Hundreds of Millions at Stake

The firm claims the designation is already “jeopardizing hundreds of millions of dollars” in business. Over the past year, the Defense Department has signed contracts worth up to $200 million each with major AI labs, including Anthropic, OpenAI, and Google.

Wedbush analyst Dan Ives warned the blacklisting may lead enterprise clients to pause Claude deployments while the case proceeds through the courts.

Anthropic CEO Dario Amodei clarified he does not oppose AI-driven weapons in principle but believes current technology lacks sufficient accuracy for fully autonomous use. He also noted the Pentagon designation has a “narrow scope” and does not affect non-Pentagon operations.

An internal memo from Amodei, published by The Information, suggested Pentagon officials were partially motivated by Anthropic’s failure to provide “dictator-style praise to Trump.” Amodei later apologized for the memo.

What Happens Next

Anthropic clarified the lawsuits do not rule out future negotiations with the government. A Pentagon spokesperson stated it would not comment on pending litigation, and a Pentagon official confirmed last week that active talks between the parties have ceased.

The second lawsuit challenges a broader supply-chain law that could extend the blacklist beyond the Pentagon to cover the entire civilian government. The scope of this designation depends on an interagency review that remains ongoing.

Shortly after Anthropic’s blacklisting, OpenAI announced a deal to provide its technology to the Pentagon’s network. CEO Sam Altman noted the Pentagon aligned with OpenAI’s principles on human oversight of weapons and opposition to mass U.S. surveillance.

Anthropic’s investors are reportedly working to limit the damage from the fallout with federal agencies.