Newspaper.fyi No. 019 · Fri, 28 Aug 2026
Technology

A judge vacated the Pentagon's Anthropic blacklist

Rita Lin said the supply-chain label was unlawful retaliation. Anthropic welcomed the ruling.

U.S. District Judge Rita Lin in San Francisco ruled on Thursday that the Pentagon's effort to blacklist Anthropic violated the First Amendment.

CNBC reported that Lin vacated the company's supply-chain risk designation and said the Defense Department also violated the Fifth Amendment's due process clause. She directed the department to rescind the guidance, directives and instructions issued against the company. "The empty invocation of national security is not a blank check to punish and retaliate against government critics," she wrote.

The designation went up in February and March after Anthropic refused to drop two limits on Claude: no lethal autonomous weapons, and no mass surveillance of Americans. Defense Secretary Pete Hegseth labelled the firm a national security supply-chain risk and said the decision was final. President Trump told agencies to stop using it.

Lin had already granted a preliminary injunction in March. At a Thursday hearing, she said she still saw no evidence that Anthropic tried to sabotage military systems, and that the government's record had "gotten worse." A Justice Department lawyer argued the issue was trust, because AI cannot be taken apart like a rifle. Anthropic's lawyer said a contract fight had been dressed up as a security threat.

An Anthropic spokesperson told CNBC: "We welcome the court's ruling that this supply chain risk designation was unlawful."

This is the first time the U.S. military has branded an American AI lab a supply-chain risk for refusing a use case.

The ruling says the government cannot punish a contractor for saying no in public. Other agencies still use Claude. The Pentagon had been winding some of that down.

The government can appeal. Watch whether the department actually pulls the designation.