A federal appeals court on Sept. 25 upheld the Pentagon's decision to exclude Anthropic from defense contracting, rejecting the AI company's legal challenge. The 2-1 ruling permits the Department of Defense to remove Claude from its supply chain and bar defense contractors from using Anthropic's products for Pentagon work.
Anthropic had contested the Pentagon's supply chain risk designation, citing restrictions it imposed on Claude's use in lethal autonomous warfare and mass domestic surveillance. The appeals court accepted the Pentagon's argument that those limitations could prevent Claude from performing requested tasks. Anthropic has also said the designation harms its reputation.
The dispute stems from a prototype agreement valued at up to $200 million that Anthropic won from the Pentagon in July 2025. The Pentagon had demanded the ability to use Claude for all lawful purposes, which Anthropic refused.
In a separate case, a California judge blocked broader government restrictions on Aug. 27, limiting the scope of the Pentagon's exclusion. The two court orders address different government actions and remain legally distinct.
Anthropic is reportedly preparing for a November initial public offering, with prospective investors anticipating a valuation around $2 trillion. The appeals court ruling introduces a specific question for prospective buyers: how much future government-related revenue Anthropic can retain given the Pentagon's exclusion and the separate California order. Nvidia has been discussing a potential anchor investment of up to $10 billion, though those talks remain preliminary and no offering price has been set.
