US Government Suspends Anthropic’s Claude Fable 5 and Mythos 5

Anthropic has suspended access to its Claude Fable 5 and Mythos 5 models following a US government directive to restrict access for foreign nationals. The company disabled both models to comply with national security orders after authorities identified a method to bypass the models’ safeguards, according to a company statement.

Why did the US government force the suspension of Claude Fable 5?

The suspension stems from a US government directive aimed at preventing foreign nationals from accessing high-capability AI. Anthropic stated that national security authorities identified a specific technique to “jailbreak” Fable 5, allowing users to bypass safety protocols.

Anthropic doesn’t agree that this is a unique failure. The company reviewed the demonstration of the technique and claimed the vulnerabilities are “relatively simple.” According to Anthropic, other publicly available AI models can discover these same vulnerabilities without needing a specialized bypass.

Despite these claims, the company disabled access for all users on Pro, Max, and Enterprise plans who were using the free preview. The Mythos 5 access program, which served vetted partners, was also shut down.

How does the “supply chain risk” designation affect AI development?

The tension between the AI industry and the state has escalated under the Trump administration. Former US Defense Secretary Pete Hegseth specifically labeled Anthropic a “supply chain risk,” suggesting the company’s tools aren’t secure enough for government use.

This designation creates a legal battleground. Anthropic is currently suing the Pentagon to challenge the label. A US judge has already ruled that the Pentagon’s directive cannot be enforced while the lawsuit is ongoing. This means government agencies and military contractors can still use Anthropic products for now.

The BBC reported it has contacted the US Department of Commerce for further comment on the Fable 5 suspension, as the government’s specific concerns remain largely classified.

What is the difference between Fable 5 and Mythos 5?

While both were suspended, the models served different purposes and user bases. According to company records, Fable 5 was designed for a broader release to Pro and Enterprise users and surpassed the capabilities of any previous public model.

Mythos 5 was more restricted. Anthropic described the Mythos architecture as “too powerful to release” without substantial safeguards. It was limited to a small group of trusted partners through “Project Glasswing,” a specialized initiative for cybersecurity research.

What happens next for Claude users?

Users currently have no timeline for the return of Fable 5 or Mythos 5. Anthropic hasn’t announced any specific modifications to the models that would satisfy government requirements.

For those needing immediate AI tools, Claude Opus 4.8 remains accessible. The ongoing lawsuit against the Pentagon will likely determine if AI labs can resist government-mandated access restrictions based on “supply chain” concerns.

Frequently Asked Questions

Are all Claude models suspended?
No. Only Fable 5 and Mythos 5 are paused. Claude Opus 4.8 and other earlier models remain available to users.

Why can’t foreign nationals use these models?
The US government issued a directive to restrict access for foreign nationals due to national security concerns and potential “jailbreak” vulnerabilities.

Is Anthropic banned from government work?
Not currently. Although labeled a “supply chain risk,” a US judge ruled that this directive cannot be enforced while Anthropic’s lawsuit against the Pentagon continues.