.png){kind=logo}
A US federal appeals court upheld the Pentagon’s decision to designate AI company Anthropic as a supply-chain risk, allowing the government to keep the company’s Claude AI models out of Pentagon systems and work performed by its contractors.
The 2-1 ruling came from the US Court of Appeals for the District of Columbia Circuit on September 25. The dispute followed Anthropic’s refusal to remove contractual restrictions on using Claude for lethal autonomous warfare and domestic surveillance. The court concluded that the Pentagon had sufficient grounds to treat Claude's continued integration into its systems as a national-security risk.
The case turned on the Federal Acquisition Supply Chain Security Act of 2018, or FASCSA. The law gives the government authority to restrict technology considered a supply-chain risk. The appeals court focused on a provision defining such a risk broadly. The majority said the language does not require the person or company creating the risk to be a foreign adversary or to have malicious intentions.
Instead, the judges found that Anthropic’s restrictions on how Claude could be used could qualify because the Pentagon feared the company might prevent the system from carrying out national-security functions that the department considered authorized and necessary.
Anthropic sought to maintain limits preventing Claude from being used for fully autonomous weapons and mass domestic surveillance. The company challenged the Pentagon’s decision, arguing that the designation was unlawful and unconstitutional.
The appeals court rejected those challenges in the FASCSA case. The majority said the government had not exceeded its authority under the relevant supply-chain provision.
However, the ruling does not establish that every AI company with ethical restrictions can automatically be blacklisted. The decision addresses Anthropic’s specific designation and the Pentagon’s concerns about Claude’s integration into its information systems.
Also Read: Australia Calls OpenAI and Anthropic CEOs to Senate AI Inquiry
Judge Karen Henderson dissented from the majority decision. She argued that the statutory language should be read in the context of actions such as deliberately disrupting, manipulating, or compromising the operation of technology already within the federal supply chain.
The ruling follows a separate August decision in which a California federal judge blocked another Pentagon designation involving Anthropic, creating two different legal tracks around the same broader dispute.
The latest decision leaves Anthropic’s FASCSA designation in place and adds a significant legal development to the continuing debate over how government agencies can regulate and procure advanced AI systems.
