The White House Wants Anthropic to Block All Jailbreaks. That May Not Be Possible
The Trump administration is demanding that Anthropic implement stricter safeguards to prevent jailbreaks in its Claude Fable 5 model. Anthropic maintains that the risks are minimal, while the NSA insists that the model's current guardrails are insufficient.
Why it matters
The dispute underscores the technical and political challenges of enforcing safety standards on advanced AI models that may be inherently difficult to secure.
Photo-illustration: WIRED Staff; Getty Images Comment Loader Save Story Save this story Comment Loader Save Story Save this story The Trump administration’s disagreement with Anthropic over its most advanced AI models appears to be fast coming to a head.
The reporting balances the government's security concerns with the company's technical defense.
Get smarter about the news
Sign up free for a feed built around what you actually care about, Dive Deeper research on any story, and the full text of every article.
Create free accountAlready have an account? Sign in