US Government Orders Anthropic to Pull AI Models Offline

The Trump administration has ordered AI company Anthropic to pull its two newest models, Fable 5 and Mythos 5, offline, citing unspecified national security concerns. The directive requires that the models not be used by any foreign nationals, a condition Anthropic stated it could not meet, leading to the complete removal of the models.

Reports indicate the White House was prompted to act after Amazon researchers allegedly discovered a method to bypass Fable 5's safety guardrails. Amazon CEO Andy Jassy reportedly communicated these concerns to the administration, which then issued the order.

The move has sparked debate among cybersecurity experts, with many signing an open letter to President Trump to revoke the order. They argue that removing advanced cybersecurity capabilities from network defenders in the U.S. is dangerous. Some analysts suggest the action may be retaliatory, stemming from a strained relationship between Anthropic and the administration, and could potentially benefit competitors by slowing Anthropic's progress.

Katie Moussouris, a cybersecurity researcher, stated that the bypass described in a paper shared by Anthropic should not have triggered an export control, calling the directive hasty and misguided. Justin Hendrix, editor of Tech Policy Press, warned that the action could raise alarms in foreign capitals about the reliability of American AI and sets a dangerous precedent for government control over software release.