Editorial illustration
A federal judge appointed by Joe Biden just decided she knows more about national security than the Pentagon, blocking the Trump administration’s move to designate AI firm Anthropic as a supply chain risk, according to court documents from the Northern District of California. The ruling represents yet another example of judicial overreach into matters that should be left to the executive branch, especially during a time of active military operations.
Judge Rita Lin issued the injunction late Thursday, pausing the administration’s broader effort to restrict Anthropic while the case proceeds, as detailed in the court’s order. She gave the government one week to appeal, which they almost certainly will. Under Secretary of War Emil Michael’s response, as quoted in Pentagon press releases, called the ruling factually flawed and warned that it seeks to upend the President’s role as Commander in Chief. But the dispute centers on a fundamental question that shouldn’t require a courtroom to answer: Can private companies dictate terms to the United States military about how their technology gets used in national defense?
Anthropic, which makes the Claude AI system currently embedded in classified military workflows, has been refusing to allow its platform for what it considers “fully autonomous weapons” or “mass surveillance of Americans,” according to statements from the company’s filings. The Pentagon maintains that such uses are already prohibited by existing policy, that humans remain in the loop for lethal decisions, and that the military doesn’t conduct domestic surveillance, as outlined in official Defense Department guidelines. They argue, correctly, that private contractors cannot unilaterally restrict how their systems are employed in lawful military operations, so the administration moved to designate them a supply chain risk, effectively banning them from defense contracts.
Judge Lin called that designation “likely both contrary to law and arbitrary and capricious,” as per her written opinion. She went further, claiming that nothing in statute supports what she called the “Orwellian notion” that an American company may be branded a potential adversary for expressing disagreement with the government. And it’s dramatic language, but it misses the point entirely. This isn’t about disagreement; it’s about a contractor trying to impose operational constraints on the military during wartime.
War Secretary Pete Hegseth has been clear about where the administration stands, according to White House statements. He described Anthropic CEO Dario Amodei as a “master class in arrogance” and a “textbook case of how not to do business with the United States Government.” The company holds a two-hundred-million-dollar Pentagon contract awarded last year, as reported in government procurement records, yet seems to believe it can pick and choose which lawful uses are acceptable.
The practical implications are significant. Anthropic’s Claude system is the only commercial AI platform approved for classified use, and it’s deeply embedded in current military workflows, per Defense Department assessments. Replacing it would take time and resources. Meanwhile, OpenAI has already secured a competing Pentagon deal, as noted in industry reports, positioning itself as the alternative as tensions with Anthropic escalated.
What’s really at stake here goes beyond one contract or one company. The ruling raises serious questions about judicial interference in national security decision-making. As one observer noted in a think tank analysis, can a judge really order the Department of War to use a vendor that the administration considers a security risk? Judge Lin tried to thread that needle, acknowledging the Pentagon can choose not to use Anthropic while still blocking the formal designation that would prevent contractors from working with them.
It feels like a distinction without a difference, and the administration clearly views it as judicial activism encroaching on executive authority. The bipartisan group of retired judges supporting the decision, as cited in legal briefs, warn about chilling effects on free speech and business activity, but that concern seems misplaced when applied to wartime contracting decisions. The administration has one week to appeal, and given the stakes for military operations and the broader principle of civilian control over defense policy, an appeal seems inevitable.
This fight is far from over, and the outcome could shape how much authority federal courts have to second-guess national security determinations for years to come. Providence watches over the bold.
About The Author
