US Commerce Department lifts export restrictions on Anthropic's AI models

The US Department of Commerce has lifted its export ban on Anthropic's AI models Mythos 5 and Fable 5, allowing the company to reinstate customer access starting July 1.

A few weeks ago, the government ordered Anthropic to suspend access for foreign nationals, even those working in the US, to its newly launched AI models. In response, the company blocked all access to ensure compliance. Now, after receiving permission from the Commerce Department, Anthropic announced on X that it will begin restoring access tomorrow.

Previously, Anthropic had received authorization to redeploy Mythos 5 to a limited set of US organizations that operate and defend critical infrastructure. This rollout gave around 100 organizations, part of the Project Glasswing initiative, access to the company's most powerful cybersecurity model. However, the fate of Fable 5, designed to make Mythos' capabilities available to the public, remained uncertain.

According to the Financial Times, Commerce Secretary Howard Lutnick informed Anthropic in a letter that the ban on Fable was lifted after the company agreed to "proactively detect and address security risks associated with the models." Earlier, the department imposed restrictions after Amazon and other companies warned that the models were vulnerable to jailbreaks and could be used for malicious purposes. Anthropic had defended its models, listing safeguards in place to prevent misuse.