It is a no-holds-barred legal battle between the Pentagon and the artificial intelligence company Anthropic.
On Tuesday, during a hearing before a Washington appeals court, the Trump administration defended its decision to classify Anthropic as a supply chain risk while actively seeking to understand how to use its most powerful model, Mythos, to counter cyber threats, reports Axios. On the other side, the company developing Claude defended itself by arguing that it cannot control its AI models if used in classified environments.
As Reuters recalls, the lawsuit is one of two filed by Anthropic against the unprecedented move by U.S. Secretary of Defense Pete Hegseth, which came after Anthropic refused to authorize the use of the AI-based chatbot Claude for U.S. surveillance or autonomous weapons, due to safety and ethical concerns.
The AI startup led by Dario Amodei has requested the suspension of the measure at the Washington Court of Appeals and has also sued the Department of War (formerly Department of Defense) in a federal court in Northern California.
On one side is Anthropic, which argues that being blacklisted is retaliation for its stance on AI safety. This label, which prevents the company from obtaining federal contracts, is usually reserved for organizations from hostile foreign countries, notes Afp. On the other side is the Department of Defense, which officially designated Anthropic as a supply chain risk in early March, stating that the company threatened national security.
As Axios points out, treating a U.S. company as a national security threat while trying to use its technology to fight foreign adversaries is a delicate situation.
All the details.
HOW THE JUDGES OF THE DISTRICT OF COLUMBIA COURT OF APPEALS ARE DIVIDED
Regarding yesterday’s hearing, each party had 15 minutes to present their case to the panel of three judges – Karen Henderson, Gregory Katsas, and Neomi Rao – of the federal Court of Appeals for the District of Columbia, who will then issue a written opinion.
The judges asked both parties questions about the court’s jurisdiction, whether Hegseth and the Department of Defense followed procedures, and why they did not adopt “less invasive” solutions.
According to Politico, judges Katsas and Rao, both appointed by President Donald Trump, expressed skepticism about the District of Columbia Court of Appeals’ authority to review the Pentagon’s designation of Anthropic as a national security threat but suggested they might review specific orders banning a company and its products from government contracts.
Judge Henderson’s position (appointed by George H.W. Bush) was firmer, stating that the Department of Defense’s actions appear to be a “stunning overreach.” “It does not seem to me that the Department has in any way supported its conclusion about the existence of a supply chain risk from Anthropic, let alone a significant risk,” Henderson added.
THE PENTAGON’S POSITION
Sharon Swingle, a Department of Justice attorney, argued that the Department of Defense needed to act quickly and that the designation served to make the entire agency aware of the risk simultaneously. “It warned everyone that they needed to move as fast as possible to integrate replacement AI models into the system instead of Anthropic’s model,” Swingle said.
ANTHROPIC’S DEFENSE
Turning to the AI company’s position, Anthropic’s lawyer, Kelly Dunbar, explained that if the government does not trust its model, it could simply decide not to do business with the company, a less invasive measure that Congress requires agencies to consider first.
According to Dunbar, blacklisting is equivalent to a permanent legal exclusion that risks leading to broader commercial exclusions at the government level and damages the company’s reputation by labeling it as a national security threat.
WHAT TO EXPECT NOW
The front involving Anthropic and the Pentagon remains open.
“A non-unanimous decision between the Washington D.C. federal Court of Appeals and a San Francisco court means Anthropic will not be able to enter into new defense contracts but can continue to operate with non-Pentagon agencies during the litigation,” reports Axios. Meanwhile, the Washington Court of Appeals, recognizing the risk of irreparable harm to the company, has decided to expedite the case review and could rule within weeks, although proceedings of this type can take longer. In the background remains the deadline set by President Trump, who gave the Pentagon until August to exclude Anthropic from its activities.
