Cleared for Takeoff: US Lifts the Two-Week Export Block on Anthropic's Most Powerful AI Models
News Summary
The United States Department of Commerce has fully lifted export controls on Anthropic's two frontier AI models — Claude Fable 5 and Claude Mythos 5 — following two weeks of intensive government review, clearing the path for broad commercial access starting July 1, 2026 (Eastern Time).
Background: A Rapid Launch and an Unexpected Halt
Anthropic publicly released Claude Fable 5 and Claude Mythos 5 on June 12, 2026 (Eastern Time), positioning them as its most capable models to date. The launches generated immediate market enthusiasm and strong developer interest. However, within seventy-two hours of their release, the U.S. government issued an emergency directive under national security and export control authorities, requiring Anthropic to suspend access for all foreign nationals — including foreign-national Anthropic employees — both inside and outside the United States.
The directive, issued by the Bureau of Industry and Security (BIS) under the Department of Commerce, effectively took both models offline globally. Anthropic complied swiftly, disabling API and consumer access to both Fable 5 and Mythos 5 across all regions.
Why the Controls Were Imposed
Reports surfaced that researchers had identified a "jailbreak" method capable of bypassing the safety guardrails built into both models, potentially enabling their use as unrestricted cybersecurity tools. Amazon CEO Andy Jassy reportedly alerted federal officials to the associated risks, accelerating the government's response. The BIS cited its authority under export control law to mandate the suspension as a precautionary measure while a formal review was conducted.
The episode highlighted a broader policy challenge: as frontier AI models reach capabilities that intersect with national security concerns — including advanced reasoning, autonomous code generation, and multi-step task execution — regulatory frameworks must keep pace with deployment timelines.
A Staged Restoration
The government took a phased approach to lifting the restrictions. Around June 26, 2026 (Eastern Time), Commerce Secretary Howard Lutnick announced that Mythos 5 had been cleared for release to a select group of U.S. companies and federal agencies, particularly those supporting critical infrastructure. This partial restoration signaled that the review was progressing constructively.
On June 30, 2026 (Eastern Time), Secretary Lutnick announced that the BIS had fully withdrawn the export controls covering both Claude Fable 5 and Claude Mythos 5. In a public statement, the Commerce Department noted that the government "worked closely with Anthropic to analyze and approve Fable 5 to ensure alignment across the U.S. Government and strengthen America's leadership in AI."
What Changes for Customers and Developers
Beginning July 1, 2026 (Eastern Time), Anthropic began restoring access to both models across its commercial customer base. Reports indicate that more than 100 companies are now cleared to integrate Fable 5 and Mythos 5 into their products and workflows. The removal of export controls also eliminates licensing requirements that had previously applied to the export, re-export, and certain transfers of the affected models, making international deployment substantially more straightforward for enterprise customers.
Anthropic confirmed the restoration in a public statement, expressing appreciation for the collaborative process with the government and reaffirming its commitment to responsible deployment of advanced AI systems.
Implications for the AI Industry
The two-week episode offers a concrete illustration of how national security considerations can directly shape the availability of cutting-edge AI models. The rapid imposition and subsequent lifting of controls demonstrated both the government's capacity to intervene decisively and its ability to work with industry to reach a resolution within a compressed timeframe.
For AI developers and enterprise customers, the situation underscores the importance of building robust compliance frameworks into model deployment pipelines. It also raises broader questions about the oversight infrastructure needed to govern increasingly capable AI systems in a way that supports innovation while addressing legitimate security considerations.
Analysts expect the government-Anthropic collaboration model established during this review period to inform future arrangements involving other frontier AI systems.