Anthropic Pulls Claude Fable 5 and Mythos 5 Globally After US Government Export Order

Anthropic has suspended all public access to its Claude Fable 5 and Claude Mythos 5 AI models, just three days after their commercial release, following an unprecedented export control directive from the Trump administration. The order, issued late last night, requires Anthropic to immediately prevent foreign nationals from using the models, citing unspecified national security authorities.

The company responded by blocking access globally, meaning no users anywhere in the world can currently reach Fable 5 or Mythos 5, including paying enterprise customers and even Anthropic employees. The move affects what the company had described as its most powerful models to date, deployed to hundreds of millions of users.

Government Cites Jailbreak Risk

The Commerce Department's directive stems from concerns that a discovered jailbreak technique could allow the models to be used for malicious purposes, potentially posing a national security threat. Anthropic publicly disagreed with the decision, stating in a blog post: "We disagree that the finding of a narrow potential jailbreak should be cause for recalling a commercial model deployed to hundreds of millions of people."

• The directive was issued under export control authorities, not a formal recall process.
• Anthropic has not disclosed the specific jailbreak method or its potential impact.
• Enterprise customers with existing contracts for Fable 5 and Mythos 5 are now unable to access the models.
• The suspension is indefinite, pending further government review or resolution.
• Anthropic's safety warnings about the models may have inadvertently triggered the government's action, according to some analysts.

Implications for AI Regulation and Enterprise Customers

The incident marks the first time a US government has ordered a major AI company to pull a commercial product from the market over national security concerns. It sets a precedent that could reshape how AI companies handle safety disclosures and export controls. For enterprises that had integrated Fable 5 or Mythos 5 into their workflows, the sudden loss of access creates immediate operational disruptions, with no clear timeline for restoration.

Anthropic has not announced any alternative models or workarounds for affected customers. The company is expected to engage with the Commerce Department to seek a resolution, but no meetings have been publicly scheduled. The broader AI industry is watching closely, as similar actions could be taken against other frontier models if vulnerabilities are identified.