A federal judge has handed Anthropic a major legal victory, ruling that the Pentagon’s campaign to blacklist the AI company was driven more by political punishment than legitimate national security concerns. The 43-page opinion by Judge Rita Lin, as detailed by MIT Tech Review, paints a picture of a government that chose social media grandstanding over proper legal channels, and paid the price in court.
The backstory matters here. The U.S. government used Anthropic’s Claude throughout much of 2025 without any complaints. Defense employees accessed it through Palantir under a government-specific usage policy that, according to Anthropic cofounder Jared Kaplan’s court declaration, “prohibited mass surveillance of Americans and lethal autonomous warfare.” Things only went sideways when the government tried to contract with Anthropic directly.
What the Judge Found
The court identified a clear pattern: officials tweeted inflammatory statements first and consulted lawyers second.
- President Trump posted on Truth Social on February 27, calling Anthropic staff “Leftwing nutjobs” and ordering every federal agency to stop using the company’s AI
- Defense Secretary Pete Hegseth followed up by threatening to label Anthropic a supply chain risk
- Hegseth also claimed that “No contractor, supplier, or partner that does business with the United States military may conduct any commercial activity with Anthropic”
The problem? The government’s own lawyers admitted in court that the Secretary doesn’t actually have the power to enforce that statement. Judge Lin noted it had “absolutely no legal effect at all.”
The “Kill Switch” Claim Collapsed
One of the government’s core arguments was that Anthropic posed a supply chain risk because it could implement a “kill switch” on its AI systems. According to MIT Tech Review, the government’s lawyers later had to admit they had no evidence to support that claim.
Labeling a company a supply chain risk also requires specific procedural steps. The judge found Hegseth never completed them. Letters sent to congressional committees claimed that less drastic measures were evaluated and deemed insufficient, but provided zero supporting details.
First Amendment Implications
What stands out here is that Judge Lin went further than just ruling on the contract dispute. She found that Anthropic had legitimate grounds to claim its First Amendment rights were violated. The government, she wrote, “set out to publicly punish Anthropic for its ‘ideology’ and ‘rhetoric,’ as well as its ‘arrogance’ for being unwilling to compromise those beliefs.”
This is significant. It sets a precedent that the government can’t use public pressure campaigns to retaliate against AI companies for their safety policies or public positions.
Why This Matters
This ruling sends a clear signal to the AI industry: safety-focused positioning won’t automatically make you a government target, at least not one that holds up in court. For AI companies navigating defense contracts while maintaining responsible use policies, the Anthropic case shows that courts will scrutinize whether government actions follow established legal processes or are politically motivated.
The irony is hard to miss. What started as an attempt to make an example of Anthropic ended up exposing the government’s own procedural failures and overreach. The Pentagon wanted a culture war. Instead, it got a 43-page lesson in how contract disputes actually work.
Full details of the ruling and timeline are available in the original MIT Tech Review report.