Three days. That's how long it took the US government to pull the plug on what Anthropic called its most capable AI model ever released to the public.
On June 9, Anthropic launched Claude Fable 5: a public version of its powerful Mythos model family, equipped with cybersecurity safeguards.
By June 12, it was gone.
The US Commerce Department issued an export-control directive ordering Anthropic to suspend access to Fable 5 and Mythos 5 for any foreign national, anywhere in the world. Since Anthropic has no way to filter users by nationality in real time, it did the only thing the order left room for: it switched both models off for everyone, everywhere.
Polymarket traders currently put 69% odds on Fable 5 returning to US customers by July 1. But nobody really knows if or when Claude Fable 5 will be restored. That is the question.
Why Did the US Government Ban Claude Fable 5?
The official reason was national security.
According to Anthropic's own statement, the government believed someone had found a way to bypass Fable 5's safety filters, a so-called jailbreak, that could expose its most powerful cybersecurity capabilities.
The catch: Anthropic says the reported technique isn't particularly alarming. The company reviewed what it believes is the underlying evidence and concluded that the level of capability shown is already available from other models on the market, including OpenAI's GPT-5.5, and is routinely used by cybersecurity professionals to defend systems.
In Anthropic's words, the evidence consisted of "asking the model to read a specific codebase and fix any software flaws," which is, essentially, what security engineers do for a living.
White House AI adviser David Sacks told a different story on X.
According to Sacks, a trusted partner flagged the jailbreak, the administration asked Anthropic to either fix it or take the model down, and Anthropic refused. Reporting by the Wall Street Journal and Semafor points to Amazon as the source of the original tip, notable because Amazon is both one of Anthropic's largest investors and a cloud infrastructure provider for the company.
This doesn't happen in a vacuum, either. Anthropic and the Trump administration have been at odds since early 2025, after Anthropic declined to let the Pentagon use its models for domestic surveillance and autonomous weapons. The Department of Defense subsequently labeled Anthropic a "supply chain risk"—a designation normally reserved for foreign adversaries. Anthropic is suing to reverse it.
What Needs to Happen for Fable 5 to Come Back?
That's the open question. Instead of drifting into an abstract timeline, the July 1 contract has locked in the highest percentage and sharpest attention on Polymarket, becoming the definitive target for traders measuring the speed of these high-stakes negotiations.
Negotiations are underway. Anthropic dispatched senior technical staff, including co-founder Tom Brown and policy chief Sarah Heck, to meet with Commerce Department officials and the National Cyber Director in Washington.
Commerce Secretary Howard Lutnick has been in regular contact with Anthropic leadership. Both Lutnick and CEO Dario Amodei are expected to attend the G7 meetings in France, where talks may continue.
Anthropic has called the situation "a misunderstanding" and says it is working to restore access as soon as possible. The problem is that no timeline has been announced. There is no deal yet, no public statement that the government's concerns have been addressed, and no clear bar for what Anthropic would need to do to satisfy the directive.
Resolution might also require technical changes: stricter filters, identity verification systems for users, or some form of nationality-based access control. Any of those takes time to build and test. Washington negotiations tend to move slowly even when both sides want a deal.
So while Polymarket signals real optimism, a fair chunk of the uncertainty remains. Governments don't typically reverse export-control directives overnight, and Anthropic still has a lawsuit open against the same administration it's now trying to negotiate with.
Ironically, just days before the shutdown, Amodei published a policy essay comparing Washington's regulatory pace to Treebeard, the wise but painfully slow-moving tree-giant from The Lord of the Rings.
He argued that government usually operates at a completely different, much slower speed than frontier tech. But this time, Treebeard moved fast to pull the plug. Whether he can move just as fast to flip the switch back on is the ultimate question the market is watching.
