Judge Rules Trump Administration Illegally Retaliated Against Anthropic Over AI Red Lines
A federal judge ruled that the Trump administration illegally retaliated against Anthropic after it refused to let Claude be used for military applications such as mass surveillance and lethal autonomous warfare. The Pentagon had demanded removal of usage limits and acceptance of “all lawful uses,” then the administration blacklisted Anthropic after public criticism, including a presidential Truth Social post and a Defense Secretary tweet. The government’s main national-security justification collapsed when it conceded Anthropic had no backdoor access to deployed models and that Claude was no riskier than other “black box” systems. The judge found the blacklisting was retaliatory, violated the First and Fifth Amendments, fell outside the relevant statute, and was arbitrary and capricious. The court vacated the designation and boycott order, issued a permanent injunction, and denied a stay. Anthropic said the restrictions could have cut defense revenue by 50% to 100%.
