Anthropic has turned off Claude Fable 5 and Claude Mythos 5 for every customer after the US government issued an export control directive targeting both models. The shutoff lands just days after the company released Fable 5 as its most capable publicly available model.
Quick answer: Fable 5 and Mythos 5 are unavailable to all users right now. Every other Anthropic model still works normally. Anthropic says it is complying with the order while disputing the reasoning and trying to restore access.
What the directive does
The directive, issued under national security authorities, suspends all access to Fable 5 and Mythos 5 by any foreign national, whether they are inside or outside the United States. That restriction even covers Anthropic’s own foreign-national employees.
Because the company cannot reliably separate foreign nationals from the rest of its user base in real time, the practical result is a hard shutoff for everyone. US customers lose access too while Anthropic ensures it is in compliance. The order also requires a license for the export, re-export, or domestic transfer of the affected models.
| Question | Status |
|---|---|
| Who is blocked | Every foreign national, inside or outside the US, including Anthropic staff |
| Net effect | Fable 5 and Mythos 5 are disabled for all customers worldwide |
| What stays up | All other Anthropic models keep running normally |
| Directive received | June 12, 2026 at 5:21pm ET |
Why the government acted
Commerce Secretary Howard Lutnick sent a letter to Anthropic chief executive Dario Amodei informing the company that Fable 5 and Mythos 5 would be subject to export controls. The letter did not spell out the specific national security concern.
The trigger appears to be a claim from another company that it had found a way to “jailbreak” Mythos, which raised alarm inside the administration about possible national security risk. The administration had previously tried to stop Anthropic from releasing the model and failed.
Anthropic disputes the basis for the order
In its public statement, Anthropic said the government has so far provided only verbal evidence of a narrow, non-universal jailbreak. The technique essentially asks the model to read a specific codebase and fix software flaws.
The company reviewed a demonstration of the technique and said it surfaced only a small number of previously known, minor vulnerabilities. Those flaws look relatively simple, and Anthropic says other publicly available models can find the same issues without any bypass. It points to OpenAI’s GPT-5.5 as an example of comparable capability that defenders already use every day.
Anthropic argues that a narrow potential jailbreak should not justify recalling a commercial model deployed to hundreds of millions of people. It warns that applying that standard across the industry would effectively halt new model deployments from every frontier provider. The company called the situation a likely misunderstanding.
How the safeguards were supposed to work
When Anthropic launched the models on June 9, 2026, it shipped Fable 5 with safeguards that route sensitive queries to Claude Opus 4.8 instead. Topics tied to cybersecurity, biology, chemistry, and distillation can trigger that fallback, which Anthropic says fires in less than 5% of sessions. Some users complained the filters were too broad.
Ahead of launch, the company red-teamed Fable’s safeguards for thousands of hours with the US government, the UK AISI, third-party organizations, and internal teams. It says testers have not found a universal jailbreak, the kind of bypass that broadly unblocks a wide range of cyber capabilities. Anthropic frames its approach as defense in depth, pairing narrow or expensive-to-produce jailbreaks with monitoring and a 30-day data retention policy meant to detect and shut down attacks quickly.
Why this matters for the industry
This is one of the first times export controls have been aimed at a deployed commercial AI model rather than at chips or hardware. For years the export-control fight centered on advanced semiconductors, so extending that regime to a live software product sets a precedent for how frontier labs ship, who they can serve, and where their engineers can be based.
The immediate fallout hits the businesses and developers who built on Fable 5. Teams outside the US lost access overnight, and US teams lost it while compliance is sorted out. The inclusion of foreign-national employees also raises hard questions for any US lab that staffs research and engineering teams globally.
What happens next
Anthropic says it is working to restore access as soon as possible and plans to share more details within 24 hours of the order. It maintains that the government should be able to block unsafe deployments through a process that is transparent, fair, and grounded in technical facts, and it argues this action did not follow those principles.
For now, the things worth watching are formal documentation from the Commerce Department, any narrowing or clarification of the directive, and how rival labs respond now that a frontier model can be pulled from the market by government order rather than by a company’s own choice.
