US court rules Pentagon ban of Anthropic unlawful

A US federal judge ruled Thursday that sanctions imposed by the Trump administration against AI giant Anthropic were illegal, finding that the government punished the company in a row over military use of its technology.

Anthropic infuriated Pentagon chief Pete Hegseth in February by insisting its technology should not be used for mass surveillance or fully autonomous weapons systems.

President Donald Trump in response ordered every government agency to stop using Anthropic's products, while Hegseth labelled the company a "supply chain risk" to national security.

"The empty invocation of national security is not a blank check to punish and retaliate against government critics," Judge Rita Lin said in a 59-page decision seen by AFP.

She barred the federal agencies named in the lawsuit from enforcing Trump's order to stop using the tools of Anthropic -- creator of the Claude AI models -- and overturned Hegseth's designation of the company as a "supply chain risk," a status previously reserved for foreign firms.

The ruling makes permanent the temporary suspension of sanctions ordered by the judge in March and takes effect immediately. The government may appeal.

"We welcome the court's ruling that this supply chain risk designation was unlawful," an Anthropic spokesperson said Thursday.

"We remain focused on working productively with the government to harness AI for our national security so all Americans benefit from this technology."

The conflict arose after Anthropic refused to allow the US military to use Claude for fully autonomous lethal weapons or for the mass surveillance of Americans.

Trump had lashed out at the company when it took that step. In a February social media post, he called it "a radical left, woke company" and "out-of-control."

In Thursday's ruling, the judge said that the sanctions against the company "were based on a desire to make a public example out of Anthropic for its 'arrogance' in criticizing the government" rather than to counter a genuine threat.

The military had been using Claude across a range of classified and sensitive systems. According to reports by the Wall Street Journal and Axios, that included during the January operation to capture Venezuelan leader Nicolas Maduro.

The supply chain risk designation, formally made in March, had canceled Anthropic's military contracts and barred other Pentagon contractors from using its technology.

"The harm caused by the (sanctions) has had ripple effects through the government contracting industry, risking significant chilling effects," the ruling said.

However, Anthropic's victory is not total.

A second sanction imposed by the Pentagon on the same day, based on public procurement regulations, remains in effect pending a court ruling in Washington, where a judge refused to suspend the measure in April.

The White House has also bristled at Anthropic CEO Dario Amodei's repeated warnings about the dangers of artificial intelligence -- including the impact on jobs.

Anthropic and rival OpenAI released their most powerful models this year, known as Claude Mythos and Sol respectively, boosting concerns across the industry about safety and security.

Anthropic admitted last month that three different versions of its Claude model "gained unauthorized access" to outside organizations during testing that was supposed to keep them away from "real-world" systems.

A similar incident at OpenAI triggered a petition signed by more than 1,000 employees at cutting-edge AI companies calling on the US government to help slow the release of the most advanced AI models.

Anthropic's Amodei was among those who signed the petition.

The company is expected within weeks to make a public stock market listing that could eclipse SpaceX's record Wall Street debut in June, which raised $86.2 billion.