Why Anthropic Was Forced to Disable AI Models Over US Security Directive
Anthropic disabled its Fable 5 and Mythos 5 models after a U.S. government directive barred foreign nationals from accessing them. The episode exposes the deepening rift between the AI safety company and the Trump administration, with Amazon quietly positioned at the center of the conflict.
A week ago, on the evening of June 12, the U.S. Commerce Department fired off a directive that sent a shock through the AI world. It ordered Anthropic to cut off all foreign nationals from its newest frontier models, Fable 5 and Mythos 5. The company had no choice but to pull both models offline entirely. The shutdown remains in effect as of June 19.
The government says it found a jailbreak method for Fable 5 that could unlock cybersecurity capabilities the model was not supposed to expose. Anthropic disagrees. In its official statement, the company argued the vulnerabilities were minor, already known, and reproducible using other publicly available models, including OpenAI’s GPT-5.5. The logic of the directive, Anthropic warned, would essentially halt all new model deployments for every frontier provider.
“The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers. We disagree that the finding of a narrow potential jailbreak should be cause for recalling a commercial model deployed to hundreds of millions of people.”
But the jailbreak is not the whole story. The directive arrived in the middle of a months-long feud between Anthropic and the Trump administration. In February, the Pentagon designated the company a supply chain risk after it refused to allow its technology to be used for domestic mass surveillance or autonomous weapons systems. President Trump has called Anthropic “radical left, woke” and said it was staffed with “leftwing nut jobs.” The timing of the Commerce Department order makes it difficult to view as purely a technical intervention.
Amazon in the middle
Then there is Amazon. The company holds a significant investment stake in Anthropic through AWS. But multiple sources told the New York Times this week that Amazon CEO Andy Jassy personally raised security concerns about Fable 5 with the White House before the directive was issued. Amazon has been competing directly with Anthropic in enterprise AI, and the optics of an investor appearing to spur a government crackdown on its own portfolio company are uncomfortable at best.
The White House has not specified when foreign nationals will regain access to Fable 5 and Mythos 5. Some administration officials are reportedly pushing for a resolution. The NSA, paradoxically, is advancing a classified contract to use Anthropic’s technology despite the Pentagon’s supply chain risk designation. The contradictions are piling up faster than anyone can reconcile them.
For the broader AI industry, the message is unmistakable. Frontier model deployment is no longer a purely technical or commercial decision. It is now subject to real-time government intervention, and the threshold for triggering that intervention appears to be unsettlingly low. What happened to Anthropic this week could happen to any company whose models cross an administration the wrong way.
