US export-control order forces Anthropic to pull Fable 5 and Mythos 5 for everyone

On June 12, 2026, Anthropic abruptly disabled its two most capable models, Claude Fable 5 and Mythos 5, for every customer. The shutoff was not a bug or a capacity problem — it was compliance with a US government order.

What happened

According to Anthropic's own statement, the company received a directive "today at 5:21pm (ET)" from the US government, citing national-security authorities, that ordered it to suspend access to Fable 5 and Mythos 5 "by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees."

Because Anthropic cannot reliably separate foreign nationals from the rest of its user base in real time, the only way to comply was a blanket shutoff. As the company put it: "The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance." Crucially, Anthropic added that "access to all other Anthropic models will not be affected" — Opus 4.8, Sonnet 4.6, and Haiku 4.5 keep running normally.

Why

Anthropic says the order traces back to "the finding of a narrow potential jailbreak" — a prompt sequence that coaxed a frontier model past its safety guardrails. The company disagrees with the response, calling it a likely "misunderstanding" and arguing that a single narrow jailbreak should not be grounds for recalling a commercial model. It says it is complying with the legal directive while "working to restore access as soon as possible." Reporting from Time and Fortune frames the move as the first time export-control authority has been used to switch off a live, general-purpose AI model in this way.

The ripple into developer tools

The shutoff landed just three days after Fable 5 went generally available inside GitHub Copilot. GitHub updated its June 9 changelog with a note: "Following Anthropic's announcement, effective today, access to Claude Fable 5 has been suspended across all GitHub Copilot experiences," adding that all other Claude models "remain available and unaffected." Any app, agent, or IDE that had pinned Fable 5 as its primary model went dark in the same instant.

Why it matters if you're building

For people learning to build with AI, this is a concrete lesson in supply-chain risk that has nothing to do with code quality. A model you depend on can vanish for reasons entirely outside your control — policy, legal, or geopolitical — with essentially no notice. The teams least affected were the ones that had already wired in fallback models or a provider-agnostic abstraction layer; their requests quietly rerouted to Opus 4.8 or another provider instead of failing.

As of mid-June, Fable 5 and Mythos 5 remained offline while Anthropic worked to reverse the directive. The episode is a reminder that "which model should I use?" now carries a second question: "what happens to my product the day that model isn't available?"