FOR IMMEDIATE RELEASE — September 8, 2026
WASHINGTON — The Free China Movement (FCM) calls on democratic governments, frontier-AI providers and cloud platforms to strengthen transparent safeguards against the targeted extraction of restricted AI capabilities after the U.S. National Security Agency, Federal Bureau of Investigation and Cybersecurity and Infrastructure Security Agency released a joint advisory alleging industrial-scale model-distillation campaigns by several mainland Chinese AI companies.
The agencies state that the campaigns systematically extract restricted proprietary functionalities and capabilities from U.S. frontier AI models for use in training competing systems. The advisory explicitly recognizes that model distillation itself is a legitimate and useful machine-learning technique. FCM therefore urges policymakers to focus on verifiable conduct rather than nationality, research method or political rhetoric.
FCM calls for five concrete steps: stronger account and API abuse detection; cross-provider sharing of technical indicators and repeat-abuse patterns; auditable standards for restricting access to sensitive model capabilities; coordinated safeguards among democratic allies; and publication of technical evidence whenever disclosure can be made without exposing sensitive intelligence or defensive capabilities.
Reuters reported that U.S. officials also assessed that the activity likely occurred with awareness of the Mainland Chinese Communist Government. FCM treats that as an attributed U.S. government assessment, not as independently established fact, and urges agencies to release additional technical evidence when doing so would not compromise legitimate intelligence sources and methods.
“National security claims are strongest when governments pair them with technical evidence, clear attribution and practical safeguards,” FCM said. “The answer is not to stigmatize legitimate AI research or every mainland Chinese developer. The answer is to document abusive conduct, protect restricted capabilities and give independent researchers enough information to scrutinize major public claims.”
FCM further urges frontier-AI companies to preserve relevant access logs, document suspected coordinated extraction campaigns, provide affected users and companies a meaningful process for contesting erroneous attribution, and coordinate with government agencies on narrowly tailored defensive measures.
The issue has broader implications for democratic governance. Frontier AI increasingly affects cybersecurity, military capability and critical infrastructure. Rules intended to protect high-risk capabilities should therefore be evidence-based, reviewable and applied consistently.
Related FreeChina.live coverage
- News: U.S. Agencies Accuse Mainland Chinese AI Firms of Industrial-Scale Distillation of American Models
- Editorial: AI Distillation Is Not the Issue — Covert Extraction of Restricted Capabilities Is
- Rapid Response: U.S. Agencies Warn of Industrial-Scale AI Model Distillation