Anthropic has suspended access to its Fable 5 and Mythos 5 models after the Trump administration demanded the company bar foreign nationals from using them. The ban, disclosed by Anthropic in a statement, came with only 90 minutes’ notice to comply. The affected models are among the startup’s most advanced.
Amazon CEO Andy Jassy voiced concerns about Anthropic’s AI models to U.S. officials before Washington’s crackdown, according to a source cited by Reuters. These talks helped trigger the export restrictions, marking an unusual intersection of private-sector advocacy and national security policy. The White House did not comment on its deliberations.
The U.S. government classified the models as posing a national security risk, leading to the immediate restriction. Bloomberg reports that the ban applies to any foreign national, effectively cutting off international access. Anthropic did not disclose which countries or entities were most affected.
For Anthropic, the order threatens its global partnerships and revenue from overseas clients. The company has invested heavily in safety research and positions itself as a responsible AI developer. The abrupt restriction may also chill investment in U.S. AI startups reliant on international talent and customers.
A cybersecurity CEO whose firm conducted research leading to the restrictions said the work was for defense purposes and added, “It’s not a jailbreak,” Fortune reports. The official rationale suggests the government sees leakage of frontier AI capabilities as akin to controlled technology exports.