The Central Cyberspace Administration of China deploys a 4-month special rectification of AI application chaos

The Central Cyberspace Administration of China issued a notice deploying a nationwide special campaign lasting 4 months to clean up and rectify AI application chaos, divided into two phases, each targeting 7 categories of prominent problems. The first phase is a special rectification of typical violations in AI application services, focusing on failure to fulfill large model filing and registration obligations as required, insufficient platform security and review filtering capabilities, security of large model training corpora, AI data poisoning, inadequate labeling of generated and synthesized content, misuse of AI technology for cyberattacks and face-swapping and voice-cloning, and inadequate security management of open-source models. The second phase focuses on chaos in AI information content, covering AI-modified classics and generation of digital garbage content, production and dissemination of false and inaccurate information, impersonation of others, violent and vulgar content, infringement of minors' rights and interests, AI-managed online troll farms, and violations by AI product services and applications. The notice requires local cyberspace administration departments to fulfill territorial management responsibilities, supervise websites and platforms to conduct self-inspection and self-correction against the rectification priorities, and improve long-term governance mechanisms.