US Government's Anthropic Models Ban Was Never About an AI Jailbreak

On Friday afternoon, the US Commerce Department sent Anthropic a letter invoking an obscure export control directive that banned non-Americans, including Anthropic's own employees, from accessing its Fable 5 and Mythos 5 models, citing unspecified national security concerns. Anthropic said it believes the letter is related to a bypass of the model's guardrails but is unsure because the letter provides no specific details. The letter has not been made public. In response, Anthropic shut down both models to all customers to ensure compliance.

Katie Moussouris, a cybersecurity veteran and founder of Luta Security, revealed in a blog post that Anthropic had shared a private copy of a security researchers' paper describing an alleged guardrail bypass in Fable 5. Moussouris stated that the bypass described in the paper "should never have triggered an export control." She criticized the directive as hasty, heavy-handed, and misguided. Moussouris and dozens of other top security researchers have called on the Trump administration to revoke the order, calling the removal of advanced cybersecurity capabilities from network defenders "dangerous."

The article suggests the administration's action may be retaliatory. Citing sources, Axios described a tense situation between Anthropic and the administration, attributing the export directive to "personality differences" rather than a technical issue. Justin Hendrix, editor of Tech Policy Press, said the move is "likely to raise alarms in foreign capitals about the reliability of American AI for critical applications." The aftermath sets a dangerous precedent about how much control the government intends to wield over the release of American-made software. The government's reasoning remains unclear, and the White House may be scrambling to undo the damage of its own making.