The U.S. government has ordered Anthropic to block foreign nationals from accessing its two newest and most powerful AI models, Claude Fable 5 and Mythos 5, setting off an urgent scramble in Washington and the security community.
According to CNBC, Anthropic received an export control directive requiring the company to suspend access to the models for any foreign national. Security Boulevard reported the tools were "abruptly disabled" shortly after their release. Tech Times noted that Anthropic sent top engineers to Washington to seek a resolution, and the BBC confirmed the company has secured a meeting with the White House after being forced to block users from the just-launched models.
According to TechCrunch, a group made up of dozens of cybersecurity experts has urged the White House to remove the restrictions, arguing the ban will limit defenders' ability to secure software and products. Gizmodo reported that cybersecurity professionals are baffled by the move, and Cybersecurity Dive and Yahoo both covered experts blasting the government over the restrictions.
The episode has also drawn broader scrutiny. Tech Policy Press raised the question of whether blocking Mythos sets a new precedent for how the U.S. government can control the distribution of frontier AI models. Meanwhile, analyst Ben Thompson, writing in Stratechery via Techmeme, argued that Anthropic's self-described commitment to safety has given the company license to aggressively pursue its business interests and even push back against the U.S. government.
If the ban holds, it would mark one of the first times Washington has used export controls to restrict a domestic AI company's own customers — a move that could reshape how cutting-edge AI tools reach global users and defenders.