Trump administration lifts export restrictions on Anthropic AI models

The Trump administration lifted export restrictions on Anthropic’s Fable 5 and Mythos 5 models Tuesday evening, allowing the company to begin restoring access to customers Wednesday. The move reversed a June 12 directive that had forced Anthropic to disable both systems worldwide.

The order, issued under national security authorities, required the company to suspend access to Fable 5 and Mythos 5 by any foreign national, including foreign national Anthropic employees, whether inside or outside the United States. The directive gave no specific details about the security concern, though Anthropic’s understanding was that officials believed they had found a way to bypass, or jailbreak, Fable 5. Anthropic had already spent thousands of hours red-teaming Fable 5 with the U.S. government, the UK AISI, private third parties and internal teams, and no tester had found a universal jailbreak.

The rollback followed a narrower change days earlier, when Commerce Secretary Howard Lutnick allowed Anthropic to release Mythos 5 to roughly 100 companies and federal agencies. In a June 30 letter, Lutnick said Anthropic no longer needed an export license after agreeing to proactively detect and address security risks, coordinate with the government on future-release protocols, and report malicious activity it identifies. Access to Fable 5 would return to general users in the United States and abroad.

In February, Donald Trump ordered U.S. agencies to stop using Anthropic technology, and Defense Secretary Pete Hegseth later said the Pentagon would designate Anthropic a supply-chain risk to national security after the company refused demands that would have let the military use its technology for all lawful purposes. Anthropic had pressed for limits blocking fully autonomous weapons and mass domestic surveillance of Americans.

Fable 5 is a consumer version of Anthropic’s more powerful Mythos model, with additional safeguards meant to reduce misuse in cyberattacks. The June rollback also followed pressure from AI companies and European governments that wanted access to Anthropic’s tools to find vulnerabilities before adversaries could exploit them.