Anthropic is redeploying its Fable 5 AI model globally. The move follows the lifting of US export controls on the model. Those controls had forced a full suspension weeks earlier. Fable 5 returns to users worldwide starting July 1.
The US government applied the controls on June 12. They required Anthropic to restrict access for all foreign nationals. Because the order took effect immediately, the company suspended access entirely. It had no reliable way to verify nationality in real time.
Over the past two weeks, we have worked closely with Anthropic to analyze and approve Fable 5 to ensure alignment across the US Government and strengthen America’s leadership in AI. https://t.co/PtVJWyQ9kH
— Howard Lutnick (@howardlutnick) June 30, 2026
The government lifted those controls as of June 30. Fable 5 now returns across the Claude Platform, Claude.ai, Claude Code, and Cowork. Anthropic also restored Mythos 5 for a set of approved US organizations. That followed government approval on June 26.
The dispute began with a reported jailbreak. Amazon researchers found a method to bypass Fable 5’s safeguards. It prompted the model to identify software vulnerabilities. In one case, it produced code showing how to exploit one flaw.
Anthropic said its testing put the finding in context. It stated that weaker models could identify the same vulnerabilities. These reportedly included Claude Opus 4.8, GPT-5.5, and Kimi K2.7. The company said the technique exposed no unique Mythos-level capabilities.
Even so, Anthropic moved to close the gap. It trained an improved safety classifier targeting the reported behavior. The company said the fix now blocks the technique in over 99% of cases. Blocked requests get rerouted to its Opus 4.8 model. Commerce Department researchers reviewed and endorsed the safeguards.
Anthropic also proposed a shared industry framework for jailbreaks. It is working with Amazon, Microsoft, and Google on the standard. The framework would score jailbreak severity on four criteria. The company said it is also deepening collaboration with the US government on frontier AI security testing.
