Anthropic Gets Green Light to Restrict AI Exports Lifted on Two Flagship Models

The U.S. government has cleared the way for Anthropic to once again offer its most advanced artificial intelligence systems—Claude Fable 5 and Mythos 5—to users around the world. The company confirmed late Tuesday that the Department of Commerce had removed export controls that had blocked foreign access to the models for nearly a month. Starting Wednesday, Anthropic says it will begin restoring access to both systems.

The reversal comes after Commerce Secretary Howard Lutnick sent a letter to Anthropic stating that an export license was no longer required. In exchange, the company agreed to take a more active role in monitoring its own technology: it will proactively detect and address security risks tied to the models, help the government set standards for future releases, and report any malicious activity it uncovers. Anthropic thanked users for their patience and said it worked closely with regulators to get the models back online.

The restrictions were imposed abruptly last month by the Trump administration, which ordered Anthropic to block all foreign nationals—including the company’s own employees—from accessing Fable 5 and Mythos 5. At the time, Anthropic said it had not been given a clear reason beyond vague national security concerns, though it believed officials were specifically worried about vulnerabilities in Fable 5. On Friday, the company also announced it had received approval to supply the models to U.S. organizations that operate critical infrastructure, and that it is continuing to coordinate with the government on broader public access.