A California judge on Thursday barred the Trump administration from placing Anthropic, the company behind the Claude AI models, on a government blacklist tied to national security. Reuters reports the judge characterized the administration’s decision as "unlawful and unfounded."
What happened?
The contested government action would have labeled Anthropic a "national-security risk" supplier. The designation followed Anthropic’s refusal to allow the U.S. military to use Claude models for surveillance purposes or in certain weapons. According to Reuters, the judge found the government’s decision legally and substantively unsupported.
The Reuters account attributes the decision to actions connected to Pete Hegseth; the report describes Hegseth as previously involved with defense and more recently connected to what the article terms a war ministry in its account of the decision-making.
Why it matters
A U.S. government "national-security risk" designation can be applied to companies whose technologies might make military systems accessible to adversaries, increasing sabotage risk. Such a label can have significant economic and reputational consequences. Anthropic’s leadership says the Pentagon’s earlier designation cost the company billions of dollars in potential revenue and inflicted notable reputational damage.
Reactions
Anthropic welcomed the court decision and said it will continue to focus on constructively cooperating with the government on national-security uses of artificial intelligence. The Trump administration did not comment on the matter, Reuters reports.
Next steps
The court order prevents the government from applying the contested blacklist designation to Anthropic while the legal proceedings continue. The case highlights tensions between government national-security concerns and technology companies’ choices about how their AI systems are allowed to be used, especially for advanced language models and applications like Claude.
We have previously published more detailed pieces on the Anthropic matter.
