Home / Reports / Claude Fable 5 Pulled

Thank you for reading this post, don't forget to subscribe!

News Report · The AI Index

Claude Fable 5: Pulled in Under a Week

On June 9, 2026, Anthropic released Claude Fable 5 and Claude Mythos 5 — the first models in a new “Mythos class” positioned a tier above Opus, and state-of-the-art on nearly every benchmark tested. Less than a week later, the U.S. government forced Anthropic to disable both, citing national-security concerns.

By The AI Index · Updated Jun 18, 2026 · 5 min read

What Fable 5 could do

Anthropic described Fable 5’s capabilities as exceeding any model it had previously made generally available — state-of-the-art on nearly all tested benchmarks, with exceptional performance in software engineering, knowledge work, vision, and scientific research. It was pitched as the most capable model for ambitious coding: large migrations, complex implementations, and multi-day autonomous sessions, able to write its own tests, implement designs with high fidelity, and use vision to check its outputs against goals. Mythos 5 was the higher, less-restricted tier; Fable 5 was the version made safe for public access.

The safeguards

Fable 5 shipped with guardrails designed for a more powerful model. For sensitive topics, queries were quietly routed to a response from Claude Opus 4.8 instead — a fallback that triggered, on average, in fewer than 5% of sessions. In the highest-risk areas — cybersecurity, biology, chemistry, and “distillation” — the model blocked responses entirely and deferred to Opus 4.8. The design reflected a now-familiar pattern in frontier AI: ship the capability, but wrap it in automated limits on the most dangerous uses.

The removal: a government export order

Within days, the U.S. Department of Commerce intervened. Commerce Secretary Howard Lutnick sent a letter to Anthropic CEO Dario Amodei stating that Mythos 5 and Fable 5 would be subject to export controls — barring access from any location outside the U.S. and from all foreign persons within it, including Anthropic’s own employees. To comply, Anthropic disabled both models for every customer. The stated trigger: the government believed it had become aware of a method of “jailbreaking” Fable 5 to identify minor software vulnerabilities. Anthropic publicly disagreed that a narrow potential jailbreak warranted recalling a commercial model already deployed to hundreds of millions of people.

“A state-of-the-art commercial model was treated like a controlled export and pulled within a week — the developer no longer decides alone.”

Why it matters

This is the clearest signal yet that frontier AI has crossed into the territory of national-security regulation. A state-of-the-art commercial model was treated like a controlled export and pulled within a week — an outcome that would have been unthinkable for earlier model generations. It sets a precedent: capability and availability are no longer decided by the developer alone, and the line between a “commercial product” and a “dual-use technology” is now being drawn by governments in real time. Crucially, no other Anthropic models were affected — Opus 4.8 and the rest of the lineup remained available.

Methodology & sources

• Launch, capabilities, pricing & safeguards — Anthropic; AWS
• Public release & positioning — TechCrunch; CNBC
• Government order & shutdown — MarkTechPost; InfoQ

Frequently asked

What is Claude Fable 5?

Claude Fable 5 is an Anthropic AI model released June 9, 2026 — the public, safeguarded version of the new “Mythos class,” positioned a tier above Opus and state-of-the-art on nearly all tested benchmarks, with a 1M-token context window and pricing of $10/$50 per million input/output tokens.

Why was Claude Fable 5 removed from the market?

Less than a week after launch, the U.S. Commerce Department placed Fable 5 and Mythos 5 under export controls over national-security concerns tied to a reported jailbreak method. To comply, Anthropic disabled both models for all customers, though it publicly disagreed that the issue warranted a recall.

Are other Claude models affected?

No. Anthropic stated that only Fable 5 and Mythos 5 were disabled; all other models, including Claude Opus 4.8, remained available and unaffected.

Related: All reports · The EU AI Act goes fully live · Frontier dual-use risks