Claude Mythos and Fable 5: The Story of Anthropic's Most Powerful Models That Lasted Only Three Days

If you've been following AI news lately, you've probably heard the names Claude Mythos and Fable 5. In June 2026, Anthropic released two highly capable models that generated both excitement and significant controversy. This article explains what these models were, why they mattered, and what happened to them.

What Exactly Was Claude Mythos and Why Was It Built?

The story begins in April 2026, when Anthropic introduced Claude Mythos Preview as a research preview model. It was part of an initiative called Project Glasswing and focused on defensive cybersecurity applications.

From the start, this model was released with significant caution. Anthropic stated that due to safety concerns, it had no immediate plans for general availability, limiting access to a small group of vetted organizations.

Fable 5: The Public Mythos-Class Model

On June 9, 2026, Anthropic released Claude Fable 5, the first publicly available model in its new "Mythos-class" tier, positioned above the Opus line in capability. Fable 5 carried the same core capabilities as Mythos but was released with substantially stronger safety layers to make it suitable for general users and developers.

The full, less-restricted version, Claude Mythos 5, remained available exclusively to Project Glasswing partners.

The Sudden Suspension: Just Three Days After Launch

That's where things took an unexpected turn. On June 12, 2026, only three days after Fable 5's public launch, the US Commerce Department issued an export control directive. Under this order, Anthropic was required to suspend access to Fable 5 and Mythos 5 for all foreign nationals, whether inside or outside the United States — a restriction that even applied to Anthropic's own foreign national employees.

”

"The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees." — Official Anthropic statement

Since Anthropic could not verify user nationality in real time, the company made the decision to disable access to both models for all customers worldwide. All other Claude models, including Opus 4.8, remained unaffected and available.

What Prompted This Order?

According to reports, the directive followed a claimed "jailbreak" method — a way of bypassing the model's safety guardrails — that a third-party company had discovered and reported to the government. The concern was that if these safeguards could be defeated, a consumer-facing product could effectively become an unrestricted cyber tool.

Anthropic characterized the finding as "narrow and non-universal," stating that the level of capability demonstrated was already widely available in other models, including OpenAI's GPT-5.5. In its statement, the company said:

”

"We believe the government should have the ability to block unsafe deployments, as part of a statutory process that is transparent, fair, clear, and grounded in technical facts. This action does not adhere to those principles."

Comparing Anthropic's Models

Unlike Fable 5 and Mythos 5, the last three models — Opus 4.8, Sonnet 4.6, and Haiku 4.5 — were unaffected by the suspension order and remain fully available.

Current Status

As of this writing, Fable 5 and Mythos 5 remain suspended. Anthropic has described the situation as a "misunderstanding" and is working to restore access. This marks the first time a US export control directive has forced the takedown of a publicly deployed frontier model — an event that has sparked significant debate about the boundary between national security and AI innovation.

Conclusion

The Mythos and Fable 5 episode served as an important reminder: no matter how carefully a powerful model is designed and safeguarded, when national security and export control authorities get involved, even a limited and disputed report can lead to the complete withdrawal of a commercial product. The event has triggered ongoing discussions about the transparency of government oversight processes for AI technology.

This article is based on official statements published by Anthropic and credible news reporting through the end of June 2026.