The U.S. government forced Anthropic to take its top AI models offline citing national security, but security experts say the move was hasty and based on a misunderstanding of how the models work.

What happened: On Friday, the U.S. Commerce Department sent Anthropic a letter invoking an export control directive that banned non-Americans, including Anthropic employees, from accessing Fable 5 and Mythos 5. Anthropic shut down both models to all customers to comply with the directive, even though the letter did not provide specific details about the alleged security concern.

Why it matters: Security researchers and experts, including Katie Moussouris of Luta Security, argue the export control was misguided. The guardrail bypass described in the paper—the apparent trigger for the ban—involves a difference in how you phrase a question to the model, not a meaningful security flaw that weakens the model. Justin Hendrix of Tech Policy Press warned the move sends a message to foreign governments that U.S. AI companies cannot be trusted to operate without government interference, potentially damaging America's reputation for AI reliability.

What to watch: The Trump administration has not confirmed its reasoning for the directive, and reports suggest the move may have stemmed from 'personality differences' between Anthropic and the administration rather than a technical issue. Moussouris and dozens of other security researchers have called on the Trump administration to revoke the export control order.