Anthropic has shut off global access to two of its most advanced AI models — Fable 5 and Mythos 5 — after the U.S. government issued an export control directive citing national security concerns.
According to Axios, Commerce Secretary Howard Lutnick told Anthropic that both models would be subject to export controls after an unnamed company claimed it had successfully jailbroken Mythos 5. The Trump administration is using the directive to block foreign governments, companies, and individuals from accessing the models.
Anthropic said it disabled access to comply with the government order, according to CNBC. AWS confirmed it has revoked access to Fable 5 and Mythos 5 for all users across all regions, while noting that all other Anthropic models remain unaffected.
The scope is sweeping: Anthropic's own announcement described the U.S. government as invoking national security authorities to suspend all access — not just restricting foreign use, but pulling the models entirely for every customer worldwide.
This is one of the most direct government interventions into commercial AI access on record, effectively treating a private company's frontier models like controlled military technology — a precedent that could reshape how advanced AI is developed, deployed, and governed globally.