# Anthropic Suspends Claude Fable 5 and Mythos 5 — US national-security order

> Anthropic must disable both flagship models after a US export-control order tied to a narrow jailbreak finding.

- Org: Anthropic
- Date: 2026-06-12
- Category: ecosystem, model
- Source: https://www.anthropic.com/news/fable-mythos-access
- Canonical page: https://ai-tldr.dev/releases/anthropic-fable-mythos-suspension-jun12/

## In short

The US government issued a national-security export-control directive on June 12 ordering Anthropic to block access to Claude Fable 5 and Mythos 5 for every customer. Anthropic complied while disagreeing with the order.

## What it is

On June 12 at 5:21 pm ET, Anthropic received a US government directive ordering it to suspend access to Claude Fable 5 and Claude Mythos 5 for any foreign national, including its own foreign employees. To comply, Anthropic disabled both models for every customer worldwide. Its other Claude models, including Claude Opus 4.8, stay available.

## How it works

The order cites national security with no public detail. Anthropic says it understands the trigger was a narrow potential jailbreak in Fable 5 — specifically when the model is asked to review a codebase for software vulnerabilities. Anthropic argues that a single narrow jailbreak should not force the recall of a model used by hundreds of millions of people, and says the process did not follow the kind of transparent, fair review it called for.

## Why it matters

Every customer calling the Fable 5 or Mythos 5 API IDs sees requests fail. Teams need to roll back to Claude Opus 4.8 or an earlier model while Anthropic works to restore access. The directive also sets a precedent for how US export controls can be applied to a live commercial AI product over a single capability concern.

## Who it's for

Anthropic API customers, Claude developers

## Links

- [Anthropic Statement](https://www.anthropic.com/news/fable-mythos-access)
- [Fable 5 / Mythos 5 Launch](https://www.anthropic.com/news/claude-fable-5-mythos-5)
- [Dario: AI Exponential](https://darioamodei.com/post/policy-on-the-ai-exponential)