Federal Judge Rules Pentagon's Actions Against AI Company Anthropic Were Illegal

Key points

• A federal judge ruled that the Pentagon's measures against AI company Anthropic were unlawful and without factual basis.
• The US Defence Department had labelled Anthropic a supply chain risk, flagging it as a potential security threat within government technology procurement.
• The ruling, first reported by SecurityWeek, came after the Pentagon applied that designation earlier this year.
• No data breach or criminal group is involved; this is a government regulatory action a court found illegal.

A federal judge has told the US Department of Defence that its actions against Anthropic, the San Francisco company behind the Claude AI assistant, were illegal and had no factual foundation. The Pentagon had classified Anthropic as a supply chain risk, a formal government determination that a vendor could introduce security problems into technology the military and federal agencies buy.

That label can effectively shut a company out of government contracts and signal to other buyers that the firm may be untrustworthy.

What did the Pentagon actually do?

The Defence Department placed Anthropic under a supply chain risk designation. It argued the company posed a potential security threat. Anthropic contested that, saying the government had no legitimate grounds for it.

The court agreed. Earlier this year we reported on how defence contractors were struggling to document their readiness for the Pentagon's security rules, and this ruling lands in that same unsettled environment.

Should this concern ordinary people?

Directly, no. No customer data was stolen and no systems were broken into. It's a legal fight between a private technology company and a government department over regulatory classification.

The case still carries weight for anyone using AI tools in a workplace that sells services to the government, or whose employer relies on Anthropic's technology. A supply chain risk label can ripple through an entire business relationship. The court has now removed that label's legal basis, at least for now.

What happens next?

This ruling doesn't end the dispute. Government legal battles move through multiple rounds of appeals, and Anthropic's legal team will use this judgment to resist further restrictions while the broader case continues.

The practical watch here is for organisations procuring AI tools for work touching federal contracts. Courts are still working out where government authority over commercial AI companies begins and ends, and this case is an early data point. The government could appeal, and the designation could return in a modified form that survives judicial scrutiny.