During a 90-minute hearing in San Francisco federal court, U.S. District Judge Rita Lin pressed the Trump administration for its rationale behind denouncing Anthropic as a security threat—actions the judge said did not appear to match the national-security concerns the government cited.
Lin’s questioning came as she weighed the merits of the Pentagon’s designation of the Silicon Valley AI company as a supply chain risk, a move Anthropic is challenging in court and seeking to undo through an emergency order. The judge indicated she was focused on whether the government acted improperly when it applied what she described as a stigma more typical of companies tied to foreign adversaries such as China or Russia.
“What is troubling to me about these these actions is they don’t seem to be tailored to the national security concerns,” Lin said, according to the Associated Press account of the hearing.
Anthropic is asking Lin to remove the label, arguing that the government’s treatment has already harmed the company’s reputation and growth prospects. The company has described the steps as part of what it called an “unlawful campaign of retaliation,” according to its arguments presented to the court.
At the start of Tuesday’s hearing, Lin also said she would not decide a broader public policy debate about AI’s role in war. “It’s a fascinating public policy debate, but it’s not my role to decide who is right in that debate,” Lin observed, the AP reported. Instead, the judge said she was examining whether the administration improperly applied the security-risk designation to Anthropic.
The dispute reflects tensions over how AI technologies can be used in military contexts, including discussions about autonomous weapons and surveillance. The AP account said the government took its “extraordinary step” after balking at Anthropic’s attempts to prevent its technology from being used in fully autonomous weapons or in surveillance of Americans, and that the underlying disagreement began with a dispute over how the technology could be used in war.
The legal fight also includes allegations about the company’s public branding and internal government use. Anthropic contends it was tarnished by a Feb. 27 statement from President Donald Trump that blasted the company as part of the “radical, woke left.” The AP account said Trump also ordered federal employees to stop using Anthropic’s technology, including its Claude chatbot.
Pentagon officials gave a six-month phase-out period for Anthropic, the AP reported, and the account said Anthropic technology is already embedded in classified military platforms that have been used in the “Iran war.” It also said filings made in the case and other documents show the administration moved away from a broader government ban.
During the hearing, Anthropic’s lawyer, Michael Mongan, argued that even if there were procedural mistakes in how the security-risk label was declared, the company had already suffered reputational harm that required court intervention. “Anthropic has suffered irreparable and mounting injuries,” Mongan told Lin, the AP reported.
On the government side, Justice Department lawyer Eric Hamilton defended the designation and pressed for judicial restraint. Hamilton maintained that Anthropic “revealed itself to be an untrustworthy and unreliable partner in recent negotiations” and argued the administration should receive “substantial deference” in determining what qualifies as a security risk, according to the AP.
Hamilton also said the Defense Department “will continue to direct its operations without tech company influence,” the AP reported. Lin did not issue a ruling Tuesday, however. She asked both sides to file additional evidence by Wednesday and signaled she would rule before the end of the week.