US judge says Pentagon retaliated against Anthropic over free speech

A federal judge in San Francisco just said what a lot of people were already thinking, which is that the US government went after Anthropic purely out of pettiness because the company spoke up against the atrocities being committed by its so-called Department of War.

The case had landed in a California court this week, as Anthropic is trying to stop a full ban on its AI models inside government systems.

Judge says government actions look like punishment against Anthropic

Judge Rita Lin said it clearly in court. “It looks like an attempt to cripple Anthropic.” She added that actions like this “of course would be a violation of the First Amendment.”

Rita also pointed out that after Anthropic made its dispute with the Pentagon public, the government had a strong reaction. In her words, it “seems to have a pretty big reaction to that.” She also said the response did not look tied to a real national security need. She told the courtroom it “doesn’t seem to be really tailored to a stated national security concern.”

Rita then said the government may have gone further than needed because it was trying to punish Anthropic. She has not made a final ruling yet, but her tone left little doubt about her concerns.

“Everyone, including Anthropic, agrees that the Department of War is free to stop using Claude and look for a more permissive AI vendor,” said Rita. “I don’t see that as being what this case is about. I see the question in this case as being a very different one, which is whether the government violated the law.”

The lawsuit came earlier this month. Anthropic is asking the court to remove its label as a supply chain risk. That label triggered the ban and blocked its models from government use. The fight is now between the company and the Trump administration. The disagreement is about how AI tools should be used in military work.

Michael Mongan, who is representing Anthropic, told the court this kind of action has never been used against a US company before. He said, “This is something that has never been done with respect to American company.” He also said the authority being used is narrow and does not fit this situation.

Pentagon demands full access as Anthropic pushes back on military use of its AI

Before all this, Anthropic was working closely with the government. It had deals with several federal agencies.

The company signed a $200 million contract with the Pentagon in July. It was also the first AI lab to run its systems inside classified networks.

Things changed in September. Talks started about putting Claude on the Defense Department’s GenAI.mil platform. That is where everything slowed down. The Pentagon wanted full access to the technology for any lawful use. Anthropic did not agree.

A government lawyer, Hamilton, spoke about that in court. He said Anthropic was not just refusing contract terms. He said:-

“Anthropic is not just acting stubbornly. It’s not just refusing to agree to contracting terms. Instead, it’s raising concerns to DOD about how DOD uses its technology in military missions.”

At the same time, Anthropic is pushing forward with its tech. Claude can now take control of a user’s computer and finish tasks. A person can send a request from their phone, and the system will handle it. It can open apps, use a browser, and fill out spreadsheets.

One demo shows a user running late. The user asks Claude to export a pitch deck as a PDF and attach it to a meeting invite. Claude completes the job on its own.

This update puts Anthropic in the race to build AI agents that work without constant input. That race picked up speed after OpenClaw went viral this year. OpenClaw connects models from OpenAI and Anthropic. People can send tasks through apps like WhatsApp or Telegram. It runs on the user’s device and can access local files.

Don’t just read crypto news. Understand it. Subscribe to our newsletter. It's free.