Federal Judge Blocks Trump Ban on Anthropic AI, Citing Weak Evidence and First Amendment Concerns

US District Judge Rita Lin in San Francisco granted Anthropic a preliminary injunction on Thursday, halting a presidential directive that ordered all federal agencies to stop using the company's Claude AI model. The ruling blocks the government from designating Anthropic as a supply chain risk while the court considers a permanent decision. According to Meritalk, Lin issued the ruling two days after lawyers for both parties appeared in court for a hearing.

The injunction is not a final verdict. A conclusive ruling could take weeks or months, per Meritalk. But the immediate effect is clear: federal contractors can continue using Anthropic's technology while the legal battle plays out. The block applies government-wide, not just to the Defense Department, and represents a significant setback for the Trump administration's efforts to cut ties with the AI company.

During the hearing, Judge Lin expressed skepticism that the government had justified its decision. According to Bloomberg, she said the administration had not presented enough evidence "really justifying" the supply chain risk designation. TechCrunch and Axios reported that Lin indicated the government's case appeared weak. The core issue is whether the ban was a legitimate security measure or retaliation for Anthropic's refusal to agree to certain contract terms.

The dispute began with stalled contract negotiations between Anthropic and the Department of Defense. Anthropic opposed the use of its AI for mass surveillance or lethal weapon decisions, Pluang reports. The company argues the blacklisting was punishment for that stance. The judge warned, as noted by Pluang, that the government's reasoning could set a dangerous precedent against contractors who disagree with the administration.

Anthropic's lawsuit frames the ban as a violation of its First Amendment rights. The company argues the Defense Department blacklisted it after contract talks collapsed, not because of any legitimate security concern. CNBC reports that Judge Lin questioned the US government about an "attempt to cripple" the AI startup during a hearing in March.

The Department of Justice pushed back forcefully. In a court filing, the DOJ urged the court to reject Anthropic's First Amendment argument, calling the Pentagon's decision "lawful and reasonable," The Hill reports. The administration argues that labeling a company a supply chain risk falls within its executive authority and does not constitute a free speech violation. This constitutional dimension raises the stakes beyond a simple contract dispute.

A separate but related development involves the Commerce Department's ban on Anthropic's most advanced models, Fable and Mythos. According to TWiT, the ban prevented all users, regardless of citizenship, from accessing these tools. Amazon, a key Anthropic partner and infrastructure provider, reportedly raised concerns about the models' capabilities.

The TWiT report notes that industry experts see this as an unprecedented move with significant implications for American AI development. The ban on specific models differs from the supply chain designation, but both actions signal a broader government effort to restrict access to Anthropic's technology. The interplay between these two restrictions complicates the legal landscape and raises questions about how the government regulates cutting-edge AI capabilities.

The Trump administration maintains that its actions are both lawful and reasonable. The DOJ's court filing, described by The Hill, insists the Pentagon's labeling of Anthropic as a supply chain risk was justified by national security considerations. The government argues that the executive branch has broad authority to manage risks in federal procurement.

The administration's position is that contract negotiations broke down because Anthropic refused to accept standard terms, not because of any retaliatory motive. The DOJ contends that companies cannot use the First Amendment to shield themselves from the consequences of refusing to cooperate with defense requirements. The court will now have to weigh whether the government's stated reasons hold up against evidence of retaliatory intent.

This case could reshape how the government interacts with AI companies. Judge Lin's warning about setting a precedent against contractors who disagree with the administration, reported by Pluang, points to the broader implications. If the government can blacklist a company after failed contract talks, it could chill the willingness of AI firms to negotiate ethical limits.

The outcome will affect not just Anthropic but any AI company doing business with the federal government. The case tests the boundaries of executive power over technology procurement and raises questions about whether political disagreements can be recast as security risks. For the AI industry, the ruling offers temporary relief but leaves the underlying questions unresolved.

The preliminary injunction is a temporary measure. Meritalk notes that a final verdict could take weeks or months. The court will now proceed to evaluate the full merits of Anthropic's claims, including the First Amendment argument and the challenge to the supply chain designation. Both sides will submit additional evidence and arguments.

In the interim, federal agencies and contractors can continue using Anthropic's technology. The case also runs parallel to the Commerce Department's separate ban on the Fable and Mythos models, which remains in effect. How the court rules on the supply chain question could influence the legality of that model-specific restriction. The AI industry is watching closely because the decision will set boundaries for government regulation of advanced AI systems.