Pentagon vs. AI Startup: Court Debates the Right to Label a Tech Company Dangerous

The latest hearing in Washington saw three judges split over whether the Defense Department can brand a private AI firm as a national‑security threat. The company, known for its chatbot Claude, argues that the Pentagon’s warning is unfounded and retaliatory. One judge said there was no proof that the AI firm creates a supply‑chain risk, calling the move an overreach. Another judge questioned how the court could doubt the Secretary’s judgment and noted that the Department’s concerns stem from possible use of AI in autonomous weapons. The startup filed two lawsuits, one in Washington and another in San Francisco, after the Pentagon labeled it a risk and a former president ordered all agencies to stop using its technology. The company says the designation harms it financially and unfairly targets a domestic business for political leverage.

Earlier this month, the appellate panel denied an injunction that would have halted the Pentagon’s actions while the case waited for a final decision. Meanwhile, a lower court in San Francisco had already sided with the company and barred the Department from issuing the risk label. In its filings, the startup warned that it cannot safely deploy its AI in classified networks. A Department lawyer countered that the model could disrupt critical military operations, potentially endangering service members. A judge who was appointed by a former president also heard the arguments, adding to the mix of perspectives on how national‑security concerns should be balanced against commercial innovation.

Actions