The U.S. Department of Commerce lifted export restrictions on Anthropic’s Claude Fable 5 and Mythos 5 AI models in June 2026 [1].
This policy shift allows Anthropic to re-enable its most advanced models on major cloud infrastructure, including AWS, Google Cloud, and Microsoft Foundry [1]. By removing these barriers, the company can broaden access for a larger set of domestic and international partners through its Glasswing program [1].
The decision marks a reversal of previous restrictions that limited the distribution of these specific models. While the Commerce Department has granted this clearance, reports from mid-June 2026 indicate conflicting views on the extent of the opening [2]. Some reports suggest Washington continues to block certain foreign access to the most powerful models for security reasons [2].
Despite those remaining concerns, the primary move allows the company to resume full commercial availability of the Fable 5 and Mythos 5 series [1]. This ensures that developers and enterprises using the three primary cloud providers can once again integrate these models into their workflows [1].
The Glasswing program serves as the primary vehicle for this expanded access. By utilizing this framework, Anthropic intends to scale its reach to international partners who were previously excluded by the export bans [1].
“The U.S. Department of Commerce lifted export restrictions on Anthropic’s Claude Fable 5 and Mythos 5 AI models.”
The lifting of these restrictions suggests a shift in the U.S. government's approach to AI proliferation, balancing national security concerns with the economic necessity of maintaining American AI leadership. By allowing these models back onto major cloud platforms, the U.S. ensures that its domestic infrastructure remains the primary hub for the world's most advanced AI development, even as it selectively manages which international partners can access the technology.
