New court filing reveals Pentagon told Anthropic the two sides were nearly aligned — a week after Trump declared the relationship kaput
TL;DR
Anthropic filed two sworn declarations with a California federal court pushing back on the Pentagon's claim that the company poses an 'unacceptable risk to national security'.
Key Points
- According to the filings, the Pentagon told Anthropic just one week after Trump publicly ended the relationship that both sides were nearly aligned on a deal.
- Anthropic argues the government's case relies on technical misunderstandings and claims that were never raised during months of negotiations.
- The case centers on whether the US government can cut Anthropic off from certain government contracts or resources.
Nauti's Take
'National security risk' is a loaded label — which is exactly why it deserves skepticism when aimed at a US AI lab that was apparently days away from a deal. The timeline is damning: one week after Trump's public break-up, the Pentagon was reportedly signaling internal alignment.
That smells less like a genuine security concern and more like a political pressure tactic. Anthropic is right to fight this in court — because if labels like these stick without substance, the entire AI industry faces a precedent of arbitrary government coercion.
Context
The case exposes a stark gap between the Trump administration's public rhetoric and the actual state of negotiations with Anthropic. If the sworn declarations hold up, the Pentagon was telling Anthropic privately 'we're almost there' while publicly labeling the company a security risk — a glaring contradiction. The outcome could set a precedent for how the US government deals with domestic AI firms that don't fully comply with its demands.
Sources