U.S. Lifts Export Controls on Anthropic's AI Models Fable 5 and Mythos 5
By Priya Anand

The United States government has lifted export-control restrictions on Anthropic's advanced AI models, Fable 5 and Mythos 5, which were previously imposed due to national security concerns. This decision, effective July 1, 2026, allows for the global release of these models.
The restrictions were initially implemented on June 12, 2026, following a security evaluation by Amazon researchers who discovered a jailbreak in Fable 5 that exposed software vulnerabilities. This led to an immediate shutdown of both models worldwide.
On June 26, 2026, the Commerce Department approved limited restoration of Mythos 5 for trusted U.S.-based organizations. By June 30, 2026, the export restrictions were officially lifted, paving the way for Fable 5's full global access and continued limited access to Mythos 5 through Anthropic's Glasswing program.
Commerce Secretary Howard Lutnick stated that the government worked closely with Anthropic to ensure the models align with U.S. security standards, reinforcing America's leadership in AI. Anthropic has committed to enhancing collaboration with government agencies to improve model safeguards and share threat intelligence.