US Agencies Issue Warning Regarding China-Based AI Companies Over Distillation
US security agencies have announced that China-based artificial intelligence companies are conducting industrial-scale distillation campaigns to steal the capabilities of US models.
The National Security Agency, the Federal Bureau of Investigation, and the Cybersecurity and Infrastructure Security Agency issued a joint warning stating that artificial intelligence firms originating from China are carrying out systematic information extraction activities targeting US models.
Joint Security Warning
The National Security Agency, the Federal Bureau of Investigation, and the Cybersecurity and Infrastructure Security Agency have published an official cybersecurity advisory report regarding the activities of China-based artificial intelligence companies.
Industrial-Scale Distillation Activities
China-based artificial intelligence companies are carrying out aggressive activities to systematically extract the proprietary functionalities and capabilities of leading US artificial intelligence models in order to train their own models.
Method for Closing the Technology Gap
Distillation, a legitimate technique in artificial intelligence research, is being utilized by Chinese companies to close the technology gap without incurring high research and development costs.
National Security Risks
These developments threaten critical infrastructure and the technology sector by enhancing military and cyberattack capabilities that could be used against the United States and its allies.
Tactic of Concealment in the Global Ecosystem
China-based firms continue their operations by spreading them across the global artificial intelligence ecosystem to evade detection and acquire the top capabilities of US models.