June 2026: US Restricts Foreign Access to Anthropic’s AI Models
The US blocked foreign entities and nationals from Anthropic’s Mythos 5 and Fable 5 AI models in June 2026.
Why it matters: These export controls affect legal compliance for firms using cross-border AI technologies, impacting national security and data governance obligations.
- On June 12, 2026, Commerce Secretary Howard Lutnick declared Mythos 5 and Fable 5 subject to US export controls.
- Anthropic disabled both models to comply with restrictions barring foreign access, including foreign nationals on US soil.
- In February 2026, President Trump ordered federal agencies to stop using Anthropic’s technology over military use refusals.
- June 2026 executive order introduced a voluntary AI risk assessment framework to address national security without strict licensing.
On June 12, 2026, US Commerce Secretary Howard Lutnick notified Anthropic CEO Dario Amodei that the company’s leading AI models, Mythos 5 and Fable 5, would be placed under US export controls. These controls prohibit access by foreign entities and foreign nationals physically present within the US, citing national security concerns. More details are reported in this Axios article.
As a result, Anthropic disabled access to both models to comply. According to AP News, the restrictions apply not only to foreign entities abroad but also to foreign nationals inside the US, preventing their use of Mythos 5 and Fable 5.
Earlier in the year, in February 2026, then-President Trump directed all federal agencies to cease using Anthropic’s AI after the company denied allowing unrestricted military and government use. This was reported by NextGov. These steps show heightened US scrutiny over AI deployment for national security purposes.
Additionally, a June 2026 executive order established a voluntary AI risk assessment framework aimed at identifying AI risks that could impact national security. White House Chief AI Adviser David Sacks emphasized this framework encourages self-assessment by AI developers without imposing stringent licensing or regulatory mandates. This voluntary framework is intended to strike a balance between national security oversight and fostering AI innovation.
For legal professionals and corporate compliance officers, these developments present clear challenges: cross-border AI use, including by foreign personnel, is now legally restricted under export regulations. Entities using Anthropic’s models must reassess compliance strategies related to export controls, data security, and national security obligations, especially in multinational operations.
By the numbers:
- June 12, 2026 — Date when export controls were applied to Mythos 5 and Fable 5
- February 2026 — Trump’s directive for federal agencies to cease Anthropic AI use
- June 2026 — Issuance of voluntary AI risk assessment framework executive order
Yes, but: The voluntary AI risk assessment framework avoids strict licensing but may increase compliance complexity without binding regulations.
What's next: Legal teams should monitor for further guidance from the Department of Commerce and potential regulatory clarifications on AI export controls.