Tech news in 3 minutes
The US government’s Anthropic models ban was never about an AI jailbreak
The U.S. Commerce Department sent Anthropic a letter on Friday afternoon invoking an obscure export control directive, forcing the AI company to pull its latest models, Fable 5 and Mythos 5, offline. The directive banned non-Americans, including Anthropic employees, from accessing the models, citing an unspecified national security concern. Anthropic believes the letter is related to a bypass of the model's guardrails, but the letter lacked specific details and has not been made public. In response, Anthropic shut down both models to all customers to comply. This unilateral action by the Trump administration, which did not require court approval, demonstrates government interference in the AI industry and warns tech companies to comply or face shutdowns. Axios reported that "personality differences" between Anthropic and the administration, rather than technical issues, led to the directive. Security researcher Katie Moussouris revealed in a blog post that Anthropic shared a private paper from Amazon researchers describing an alleged guardrail bypass in Fable 5. Moussouris stated the bypass should never have triggered export controls, calling the directive hasty, heavy-handed, and misguided. She and dozens of other security researchers urged the Trump administration to revoke the order, warning that pulling cybersecurity capabilities from network defenders is dangerous. Past administrations have made broad export law decisions, but this directive appears retaliatory. Justin Hendrix of Tech Policy Press said the move "likely to raise alarms in foreign capitals about the reliability of American AI." The administration has not confirmed why it invoked the directive. Possible reasons include misreading the report, Amazon CEO Andy Jassy's influence, or pressure on Anthropic. The government has set a dangerous precedent for controlling American-made software releases, and the same could happen to any company in the future.