US Commerce Department Lifts Export Controls on Anthropic's Fable and Mythos AI Models

Export Controls Lifted After Three Week Suspension

Anthropic said on 30 June 2026 that the US Commerce Department has lifted export controls on its Fable and Mythos AI models, less than three weeks after the company was ordered to suspend access over national security concerns. The decision follows a period of heightened scrutiny by Washington over how advanced AI models could be misused by military or intelligence actors in countries the US considers adversarial.

Anthropic had disabled Mythos 5 and Fable 5 following a Commerce Department order issued on 12 June. Access to Mythos 5 was partially restored the following week for a small group of vetted US organisations, before the full lifting of controls was confirmed this week. A letter from US Commerce Secretary Howard Lutnick said the department reserves the right to reimpose a licence requirement should Anthropic fail to meet its commitments.

What Enterprise Buyers in the Gulf Should Note

For enterprise technology leaders across the GCC evaluating frontier AI models for regulated sectors, the episode is a reminder that access to the most capable models can shift quickly on national security grounds, independent of commercial availability or pricing. Anthropic has agreed to proactively detect and address security risks tied to its models and to work with the US government on protocols governing Mythos, Fable, and future releases. This mirrors the kind of enterprise AI governance conversations already under way between Gulf regulators and technology providers, including through initiatives such as the US UAE AI acceleration partnership.

Fable and Mythos share the same underlying model, though Fable is designed for broader public availability while some safeguards are relaxed for Mythos. Anthropic has also added a new safeguard targeting a technique it believes the government viewed as a way to bypass Fable's safety controls, though the company said the underlying vulnerabilities involved had already been identified and patched.

Part of a Broader Pattern of AI Export Scrutiny

The decision comes as the US government tightens its review of frontier AI releases more broadly. A recent executive order established a voluntary framework requiring developers to offer covered frontier models to the government for review before releasing them to trusted partners. OpenAI has separately delayed the full public launch of its next model at the government's request. Industry observers have questioned how consistently the administration applies its criteria for which organisations qualify as trusted, and how equivalent scrutiny will be applied to AI models developed outside the US.