An emergency U.S. export control directive has forced Anthropic to suspend global access to its new models after Amazon researchers discovered a critical security bypass.
The U.S. government issued an emergency export control directive against Anthropic following the discovery of a safeguard bypass in its new Fable 5 and Mythos 5 models by researchers at Amazon. The directive ordered Anthropic to block all foreign nationals from accessing the models. Unable to immediately verify the nationality of its global user base, Anthropic suspended access to both models worldwide. This escalation follows a prior conflict where the Pentagon blacklisted Anthropic after it refused to allow Claude to be used for military operations and domestic surveillance.
When commercial AI models are treated as critical national security assets, developers lose sovereignty over their weights.
- –The use of emergency export control directives on commercial software weights sets a precedent for direct federal control over AI.
- –Anthropic's refusal of the Pentagon's terms highlights the friction between developer safety guidelines and state defense requirements.
- –Amazon's dual role as Anthropic's cloud host and the security whistle-blower highlights the vulnerability of AI startups on tech giants.
DISCOVERED
52d ago
2026-06-15
PUBLISHED
52d ago
2026-06-15
RELEVANCE
AUTHOR
DianaDimit31323