WASHINGTON, United States: A federal judge struck down the Pentagon’s blocklist of Anthropic on Thursday, ruling that the Department of Defence acted unlawfully in labelling the AI company a supply-chain risk.
US District Judge Rita Lin said the Pentagon could not use national security concerns to punish government critics.
She also noted that the Pentagon continued seeking work with Anthropic after applying the risk label. Therefore, Lin said its actions did not support the claim that Anthropic posed a real software security threat.
Anthropic welcomed the decision. The company said it still wants to work with the US government on artificial intelligence for national security.
The dispute began after the Defence Department sought to use Anthropic’s Claude AI models for all lawful purposes. That included sensitive military and intelligence work.
Read: Meta’s $2 Billion Manus AI Deal Faces Scrutiny From Chinese Regulators
However, Anthropic opposed two uses. It barred Claude from mass surveillance of Americans and fully autonomous weapons.
The disagreement then grew after the Pentagon labelled Anthropic a national security and supply-chain risk. As a result, the company sued the government.
Meanwhile, the government may appeal Lin’s ruling. Anthropic is also fighting a separate Pentagon decision in the US Court of Appeals for the D.C. Circuit.
