China's CAC Official: AI Faces Five Key Security Risks, Including Misuse and Tech Hegemony

Artificial intelligence in China faces five types of security risks, according to Wang Lihong, deputy director and first-level inspector of the Cybersecurity Coordination Bureau at the Cyberspace Administration of China. Wang said technical vulnerabilities inherent to AI systems are hard to eliminate, which undermines output stability and reliability. She also pointed to the risk of extreme loss of control as model capabilities leap forward and overturn traditional security paradigms. Rapidly evolving AI applications are meanwhile accelerating the iteration of security risk patterns, while misuse, abuse and malicious use threaten social order and ethical boundaries. At the global level, Wang noted, AI development is confronting the risk of technological hegemony. The remarks were reported by China Central Television (CCTV).