Back to advisories

China-Based Artificial Intelligence Companies Conducting Industrial-Scale Distillation Campaigns Against U.S. AI Companies

Chinese AI companies are systematically extracting proprietary model capabilities from US AI services through large-scale knowledge distillation attacks using proxy networks to bypass access controls. This affects any organization exposing AI APIs or models and risks accelerated capability theft by state-aligned actors. The campaigns violate terms of service and likely have Chinese government awareness.

HIGHAdvisorySep 10, 2026
Action required
Immediately audit API access logs and model query patterns for anomalous behavior: high-volume requests, systematic output extraction attempts, or proxy/VPN traffic. Cross-reference with known Chinese IP ranges and proxy services. Implement rate limiting, query monitoring, and IP-based access controls on all AI model endpoints.
Affected products
NSACISAFBIClaudeGPT