US Agencies Warn China Systematically Extracting Frontier AI Capabilities via Distillation

US intelligence and cybersecurity agencies have issued a joint warning that Chinese actors are systematically using model distillation to extract frontier AI capabilities, including outputs and reasoning processes, to train rival models. The claim is based on official US assessments, though specific targets and scale remain undisclosed. This signals an escalation in AI technology competition and raises concerns over intellectual property and national security.