A federal judge has ruled that the Trump administration illegally blacklisted AI company Anthropic after it refused to permit its technology to be used for lethal autonomous weapons and mass surveillance of Americans. Judge Rita Lin found the measures amounted to unlawful First Amendment retaliation, rejecting the government’s national-security justification and ordering the directives to be withdrawn.

The administration had ordered federal agencies to stop using Anthropic’s products permanently and barred defence contractors from doing business with the firm, including on non-military work. Lin said the government’s risk assessment was weak, noting its concession that Anthropic lacked backdoor access to deployed systems and that its models were no riskier than other opaque AI systems; Anthropic sued in March and had already secured a preliminary injunction.