Analysis
A federal judge ruled Thursday that the Pentagon acted illegally when it blacklisted Anthropic as a national-security supply-chain risk, delivering the AI company a significant win in a legal fight that traces back to a February confrontation with the Trump administration. U.S. District Judge Rita Lin, in a 59-page order, called the Pentagon's measures against Anthropic “illegal and baseless,” NBC News reported.
How the Dispute Started
The fight began in February 2026, when President Trump and Defense Secretary Pete Hegseth accused Anthropic of endangering national security and designated the company a supply-chain risk. The trigger, according to court filings, was Anthropic's refusal to let the Pentagon deploy Claude for U.S. domestic surveillance or for autonomous weapons systems that select and engage targets without a human in the loop -- policy lines Anthropic has held publicly since well before the dispute became a legal fight. That refusal, not a technical or security failure, is what the administration cited when it moved to block Anthropic from certain military contracts.
“The trigger, according to court filings, was Anthropic's refusal to let the Pentagon deploy Claude for U.S.”
Lin's order rejected that justification outright. She wrote that neither the Constitution nor the federal statute the government invoked allows officials to “impose sweeping penalties based principally on Anthropic's critique of the Administration's views,” adding that “the empty invocation of national security is not a blank check to punish and retaliate against government critics,” ABC News reported.
A Company Racing Toward an IPO
Anthropic, founded in 2021 by former OpenAI researchers Dario and Daniela Amodei, has built Claude into one of the two dominant frontier AI model families alongside OpenAI's GPT line, with Google DeepMind's Gemini as the third major competitor. Pulse has tracked Anthropic's rapid scale-up through a string of multibillion-dollar compute commitments and a valuation that has climbed toward $965 billion ahead of a widely anticipated IPO expected as soon as October. A federal blacklist barring Anthropic from military and, potentially, civilian government contracts would have been a material business risk heading into that listing -- the kind of contingent liability an S-1 would have had to disclose in detail.
The ruling doesn't just clear Anthropic's path to Pentagon work; it draws a legal line other AI labs will now cite. OpenAI, Google DeepMind, Palantir and Scale AI all compete for the same category of defense and intelligence contracts, and several have taken varying public positions on autonomous-weapons use. Lin's order effectively tells the administration it cannot use national-security designations as leverage against a company specifically because that company drew its own red lines on military use of its technology -- a precedent competitors pursuing similar policies will now be able to point to.
What the Ruling Doesn't Resolve
The win is real but partial. The government is expected to appeal, which could tie the underlying legal question up for months. Anthropic also still faces a second, separate lawsuit in Washington, D.C. challenging a related Pentagon supply-chain risk designation -- one that could exclude Anthropic from a broader set of civilian federal contracts even if the military-specific blacklist stays blocked. Neither the appeal nor the second case is resolved by Thursday's order, and a higher court could still take a different view of how much deference national-security designations deserve.
For founders and investors building in AI-and-defense, the case is now a template worth studying regardless of the eventual appeal outcome: a company that took a clear, public stand on how its technology could be used, got penalized for it, and successfully argued in federal court that the penalty was retaliatory rather than security-driven. That's a fact pattern any AI company weighing its own red lines with government customers will want its counsel to have read closely.
What comes next is the appeal timeline and whether the D.C. case follows the same reasoning Lin applied here -- if it does, Anthropic clears both fronts before its IPO roadshow; if it doesn't, the S-1 will need to explain why one court saw retaliation and another didn't.
