China removes 39,000 accounts for AI-generated harmful content
China's Cyberspace Administration removed approximately 39,000 social media accounts during a Spring Festival campaign against AI-generated content. The accounts had used AI tools to automatically generate and publish low-quality and misleading content, including fabricated news and disturbing imagery targeting minors.
What happened
China's Cyberspace Administration deployed a 'Clear 2026' campaign during the Spring Festival to address network ecosystem problems. The campaign resulted in the removal of approximately 39,000 accounts and 708,000 pieces of illegal content across multiple platforms including Weibo, WeChat, Douyin, Xiaohongshu and Kuaishou.
Platforms were directed to address accounts using AI to generate and automatically publish low-quality content, including incoherent messages described as 'digital swill' containing confusing narratives and highly similar posts. AI-generated content also included modified versions of classic literary works containing graphic and disturbing imagery labelled as children's content.
Reported impact
The AI-generated content interfered with the network ecosystem and included misinformation such as fabricated reports of overcrowded trains and highway tunnel explosions. Some AI-generated content targeted minors by distributing disturbing imagery under children's content labels.
Accounts also used AI automation to simulate human interaction in comments, private messages and group chats, undermining authenticity of user engagement on the platforms.
Index analysis
The scale of account removals indicates widespread use of AI-automation tools to generate low-quality content at scale across Chinese social media platforms. The specific targeting during Spring Festival suggests coordination with platform governance policies.
Governance controls
Xiaohongshu announced a governance policy prohibiting the use of technical means to simulate real users, generate non-authentic content, or conduct false interactions. The platform announced graduated enforcement measures including warnings and content distribution restrictions for occasional AI-assisted account activity, and account suspension for accounts directly registered through AI-managed tools or with all public posts AI-generated.
The Cyberspace Administration stated it would maintain strict enforcement against violations and encouraged public participation in reporting violations to maintain network order.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.