If you’ve been wondering why Claude Fable 5 is down, the answer has little to do with a typical server outage. On June 12, 2026, the US government ordered Anthropic to suspend access to its two most powerful AI models—Claude Fable 5 and Claude Mythos 5—for every customer worldwide. Anthropic complied the same day, and that directive is the reason the models have vanished from the picker.

For anyone hitting an error that reads something like “There’s an issue with the selected model (claude-fable-5),” the takeaway is simple: nothing is broken on your end.

What Actually Happened

This wasn’t a capacity problem or a billing window closing early. The US government issued an unprecedented export control directive ordering Anthropic to immediately suspend all access to its top-tier Fable 5 and Mythos 5 models for foreign nationals, citing national security authorities.

The timeline moved fast. Both models launched publicly on June 9, and the directive arrived on June 12 at 5:21 PM Eastern. That meant they were pulled barely three days after going live.

Crucially, the order’s reach is what forced such a sweeping response. Because the directive applies to foreign nationals everywhere—including Anthropic’s own employees—the company says the only way to comply was to turn the models off for everyone. Anthropic could not reliably verify citizenship in real time, which is why the suspension became global rather than limited to non-US users.

Who’s Affected—and Who Isn’t

The good news for most users is that the disruption is narrow. Every other Anthropic model is unaffected. Claude Opus 4.8, Sonnet 4.6, Haiku, and the rest are running normally, so if your workflow didn’t specifically rely on Fable 5 or Mythos 5, you likely won’t notice anything at all.

Here’s how the fallout breaks down:

• No users around the world can currently access Fable 5 or Mythos 5, including paying enterprise customers and Anthropic employees internally
• Existing sessions end in errors, and new queries are automatically routed to older models like Opus 4.8
• All other Claude models remain fully operational

In short, this is a targeted model recall, not a wider Claude outage.

The Alleged “Jailbreak” at the Center of It

So why did the government step in? The concern revolves around a claimed security vulnerability. Anthropic’s understanding is that the government believes it has found a method of bypassing, or “jailbreaking,” Fable 5.

The company has pushed back on how serious that threat really is. Anthropic says the supposed jailbreak essentially consists of asking the model to read a specific codebase and fix any software flaws, and that a demonstration of the technique only surfaced a small number of previously known, minor vulnerabilities. The company further argued that these flaws appear relatively simple and that other publicly available models can find them as well, without requiring any bypass.

Notably, Anthropic says the evidence it has received so far has been limited—describing the only proof it’s been given as verbal.

Anthropic’s Response

Anthropic has made clear it disagrees with the order even as it follows it. In a blog post, the company called the situation a misunderstanding and said it is working to restore access as soon as possible, apologizing to its customers.

The company has also defended its broader safety approach. It points to a mandatory 30-day data retention policy that lets it study and shut down novel attacks, even for customers who normally hold zero-retention agreements. Anthropic’s underlying argument is that a model this capable was going to exist regardless, and that shipping it with hard guardrails, fallback routing, and monitoring is safer than a competitor releasing something equally powerful with no protections at all. By its reasoning, removing Fable 5 doesn’t remove the underlying capability from the world, since it already exists in other commercial models.

The Industry Reaction

The shutdown has stirred a wider debate about how vulnerable AI access is to sudden government intervention. One prominent AI founder called the move a “wakeup call,” urging developers to run local models on their own hardware to insulate themselves from regulatory volatility—arguing that no company or government can ever take away locally hosted models.

Competitors wasted no time seizing the moment. Chinese open-source provider MiniMax quickly promoted the open-weights availability of its frontier-class M3 model, contrasting its decentralized approach with the sudden lockout.

When Will Fable 5 Come Back?

This is the question on everyone’s mind—and there’s no firm answer yet. Anthropic has not published a return date, saying only that it believes the situation is a misunderstanding and is working to restore access as soon as possible.

Until that changes, the sensible move is to plan around the alternatives. The responsible assumption for now is that Fable 5 and Mythos 5 are down indefinitely, and users should rely on Opus 4.8 or another model in the meantime.

For most people, the practical impact is minimal—Claude itself remains very much online. But the episode has become something larger than a temporary model recall: a real-world test of how AI companies, governments, and the push toward ever-more-capable systems will navigate the tensions between capability, security, and access in the months ahead.