Anthropic AI export ban partially lifted in US, Canada regains access

The Trump administration announced on June 26 that it was partially rescinding the Commerce Department's export restriction on Anthropic’s AI models. Access to the Mythos 5 system was restored for roughly 100 vetted U.S. firms and federal agencies, while the more restricted Fable 5 model remained blocked. The original ban, imposed on June 12, was triggered after researchers reported a jailbreak that could reveal software vulnerabilities.

Following the partial thaw, the U.S. lifted the export controls on both Mythos 5 and Fable 5, allowing Anthropic to resume rolling out the models globally. Canadian organizations, including the Canadian Centre for Cyber Security, regained access to Mythos 5, and Anthropic reinstated a safety classifier for Fable 5 to block potentially risky requests. The company said it will continue coordinating with the U.S. government to expand access under its Glasswing program.