A federal judge appears poised to rule in Anthropic's favor for a second time in its legal battle with the Trump administration over a governmentwide ban on Claude, the company's AI assistant. During a Thursday hearing, Judge Rita Lin of the U.S. District Court for the Northern District of California suggested the government's position has actually deteriorated since she issued a preliminary injunction blocking the ban in March.

What's Behind the Government's Case Against Anthropic?

The dispute began earlier this year when Anthropic publicly criticized the Pentagon over how its technology was being used. In response, President Donald Trump and Defense Secretary Pete Hegseth issued directives banning federal agency use of Claude and designating Anthropic as a supply chain risk. The government's core argument centers on concerns about the company's trustworthiness and potential security vulnerabilities.

However, Judge Lin expressed skepticism about the government's reasoning. She noted that the record did not show evidence that Anthropic could alter the model after providing access, activate a "kill switch" to disable it, or that the model was delivered with hidden controls. Instead, Lin observed that the government's lack of trust appeared to stem primarily from Anthropic's public criticism of the Department of Defense.

"If that's really what the government's position is, I find it to be really troubling," stated Judge Rita Lin.

Judge Rita Lin, U.S. District Court for the Northern District of California

Lin's concern highlights a troubling implication: if the government can retaliate against a defense contractor simply for public criticism by stripping contracts and labeling the company an enemy of the state, it raises serious questions about free speech and due process protections for private companies.

How Is the Government Currently Using Claude Despite the Ban?

The situation on the ground remains complex. While the preliminary injunction temporarily blocked the ban, allowing agencies to continue using Anthropic's tools, the government has taken steps to phase out Claude in certain areas. The Department of Defense is working to complete its transition away from Anthropic systems by September 30.

At the same time, federal agencies are using Claude in various ways that complicate the picture:

• Pilot Programs: Several agencies are experimenting with Claude through sandboxed environments and pilot programs, including a General Services Administration initiative called OneGov deals that was set to end on August 30
• Ongoing Negotiations: Some federal agencies continue pursuing new contracts with Anthropic, even as others have paused negotiations on potential deals
• National Security Expansion: The government has reportedly expanded its use of Anthropic's newer Mythos model, which raised questions during the hearing about whether such expansion contradicts the government's stated position against the company

Judge Lin specifically asked the government to provide details about expanded use of Anthropic tools in national security contexts, including deployment of Mythos. A Justice Department lawyer initially argued the question involved nonpublic information, but later indicated the government might provide additional details after the hearing.

What Does This Mean for Anthropic's Future?

The judge's comments suggest Anthropic has a strong position heading into the final ruling. Lin noted that "if anything, it seems like the record, in some ways, has gotten worse for the government" compared to when she issued the preliminary injunction in March. This language typically signals a judge's inclination to rule against the losing party.

Lin

The case represents a significant moment for AI regulation and government-industry relations. It tests whether the government can use national security designations as a tool to punish companies for public criticism, or whether such actions require concrete evidence of actual security risks. The outcome could shape how federal agencies interact with AI providers going forward and establish precedent for how disputes between the government and technology companies are resolved.

Anthropic's legal team, represented by lawyers from Wilmer Hale, has argued that the company is not aware of agencies canceling contracts since the preliminary injunction took effect, though one Pentagon contract was canceled before the court's injunctive relief. The company has also worked with the Trump administration to establish guardrails around its newer Mythos and Fable models after the government initially restricted them citing national security concerns, suggesting a path toward resolution exists.