Judge voids Pentagon’s Anthropic blacklist

Anthropic, maker of Claude, refused to lift limits that would have allowed the Pentagon to use its model for mass surveillance of Americans and fully autonomous weapons, while saying it still wanted to work with the military. After negotiations broke down, the administration labeled the company a national-security supply-chain risk—a designation that can shut a vendor out of federal systems and contracts. U.S. District Judge Rita Lin ruled that the label was unlawful retaliation for Anthropic’s stance and that the government denied the company the process it was owed, blocking enforcement of the blacklist. The decision says the government may choose another AI supplier, but it cannot use a security exclusion as punishment when a contractor sets limits on how its technology will be used.