Anthropic Shuts Down Claude Fable 5 and Mythos 5 — What Automation Users Need to Know

Anthropic's most powerful models are offline after a US government export control directive. If you run automations on Claude Fable 5 or Mythos 5, here is what just happened and what to switch to right now.

What happened

On June 9, Anthropic launched Claude Fable 5 — the first publicly available Mythos-class model. It was the most capable model Anthropic had ever released to the public, scoring 80.3% on SWE-Bench Pro (11 points ahead of Opus 4.8) and powering everything from coding agents to 3D world-building.

Three days later, it was gone.

On June 12, the US government issued an export control directive suspending access to Fable 5 and Mythos 5 for foreign nationals. Because Anthropic has foreign-national employees and the order was broad in scope, the company made the call to disable both models for ALL customers globally — not just foreign users.

AWS followed on June 13, telling Anthropic it had to revoke access to the models across all regions. Google Vertex and Microsoft Foundry users were also affected.

Why the government stepped in

The government's letter did not spell out the specific national security concern. But reporting from multiple sources suggests it centered on a narrow potential jailbreak of Fable 5 — specifically, that the model's cybersecurity capabilities could be exploited for offensive purposes.

This is notable because Fable 5 was already shipped with safety classifiers. Anthropic had specifically split the Mythos model into two: Fable 5 (public, with safeguards) and Mythos 5 (same model weights, guardrails lifted, restricted to vetted cybersecurity partners through Project Glasswing).

Even with those safeguards in place, the government determined the risk was broad enough to warrant pulling the plug on everyone.

What this means for your automation stack

If you run automations that call Claude Fable 5 or Mythos 5 — directly through the API, through Claude Code, or through any MCP/agent framework — those workflows are currently broken. Here is what you should check:

• Claude Code users: If your Claude Code sessions default to Fable 5, they will fall back to Opus 4.8 automatically. Your automations will still work but performance on complex coding tasks may drop.
• API users: Any API calls using claude-fable-5 or claude-mythos-5 will return errors. Switch to claude-opus-4-8 or sonnet-4 immediately.
• AWS Bedrock / Vertex AI users: These models are no longer available in any region. Update your model routing to use Opus 4.8 equivalents.
• MCP server users: If your MCP server connects to Claude via the API, check your model ID config and update the fallback chain.

What to use instead

For most automation workflows — CRM sequences, email follow-ups, booking bots, WhatsApp auto-replies — Claude Sonnet 4 or Opus 4.8 will handle everything you need. The real-world difference for these tasks is minimal compared to Fable 5.

The bigger picture

This is the first time the US government has forced a frontier AI company to globally disable access to a model that was already publicly available. It sets a precedent: model availability is no longer just a product decision — it is a geopolitical one.

For automation agencies and freelancers, the takeaway is clear: do not build critical workflows on a single model. Always have fallback chains. Always test alternatives. The model that is cutting-edge today can be offline tomorrow.

At Systrify, we design every client automation with multi-model fallbacks built in from day one. If you want a review of your current setup to make sure a single model going down does not break your pipeline, book a free audit.