U.S. Restricts Anthropic AI Models, Sparks Silicon Valley Fears Over Foreign Talent Access

The U.S. Commerce Department warned AI startup Anthropic on June 12, 2026, about providing access to its "Fable" and "Mythos" models to "foreign persons," including foreign employees, without a license. This action, citing export controls, led to the models' removal and sparked concerns in Silicon Valley about the "weaponization" of regulation to limit AI companies' reliance on international talent. Anthropic met with government officials on June 15, 2026, to discuss re-listing the models. The Commerce Department's letter invoked Export Administration Regulations Section 744.22(b). The move followed an Amazon.com Inc. report detailing security vulnerabilities in Anthropic's models, which the company claims are "simple and common." Experts warn this could establish an ad hoc AI model licensing system, threatening U.S. AI leadership, which relies heavily on foreign talent. Data shows 38% of top 2024 AI academic paper authors completed university in China, with nearly 75% now working in U.S. firms. OpenAI, while supporting foreign talent, advised employees against coordinating with Anthropic due to antitrust compliance.