US Government Directs Anthropic to Suspend Access to Fable 5 and Mythos 5

Anthropic halts access to its latest models, Fable 5 and Mythos 5, following a US government directive citing national security concerns over potential jailbreak methods.

by HowAIWorks Team
AnthropicFable 5Mythos 5AI PolicyRegulationSecurity

Introduction

On June 12, 2026, the US government issued an export control directive requiring Anthropic to suspend all access to its most advanced models, Fable 5 and Mythos 5, for any foreign national. To ensure compliance with the order, Anthropic has abruptly disabled these models for all customers globally.

Access to all other Anthropic models remains unaffected by this directive.

The Government Directive

Anthropic received the directive late in the afternoon (5:21pm ET). While the government's letter lacked specific details, Anthropic understands the concern stems from a potential method to bypass or "jailbreak" Fable 5's safety measures.

According to Anthropic, they reviewed a demonstration of this technique, which appeared to identify a small number of minor, previously known vulnerabilities. Anthropic noted that these vulnerabilities are relatively simple and that other publicly available models can discover them without needing a bypass.

Anthropic's Defense in Depth Strategy

Anthropic maintains that Fable 5 underwent thousands of hours of rigorous red-teaming in collaboration with the US government, the UK AISI, and third-party organizations prior to launch.

The company acknowledges that perfect jailbreak resistance is not currently possible for any model. As a result, Anthropic employs a "defense in depth" strategy:

  • Ensuring jailbreaks are either very narrow or extremely expensive to produce.
  • Implementing thorough monitoring to detect and shut down attacks quickly.
  • Requiring a 30-day retention of customer data for Fable to research and mitigate threats.

Anthropic states that no testers have found a "universal jailbreak"—a method that broadly bypasses safeguards to unlock a wide range of cyber capabilities. They assert that the potential jailbreak shared with the government is a narrow, non-universal technique involving asking the model to read a specific codebase and fix flaws, a capability widely available in other models like OpenAI's GPT-5.5.

Disagreement with the Directive

While Anthropic is legally complying and removing access for all users, they have voiced strong disagreement with the action. They argue that recalling a commercial model deployed to millions based on a narrow potential jailbreak sets a dangerous precedent.

"If this standard was applied across the industry, we believe it would essentially halt all new model deployments for all frontier model providers," Anthropic stated. They reiterated their stance that government interventions should follow a transparent, fair, and technically grounded statutory process.

Next Steps

Anthropic has apologized to its customers for the disruption, calling the situation a "misunderstanding." The company is currently working to restore access to Fable 5 and Mythos 5 as soon as possible and promises to share more details in the coming days.

Sources

  • Anthropic Official Statement (June 12, 2026)

Frequently Asked Questions

Anthropic received an export control directive from the US government citing national security concerns related to a potential narrow jailbreak method.
No, access to all other Anthropic models remains unaffected. Only Fable 5 and Mythos 5 are suspended.
Anthropic is complying but disagrees that a narrow potential jailbreak warrants recalling a widely deployed commercial model, arguing it harms the industry.

Continue Your AI Journey

Explore our lessons and glossary to deepen your understanding.