Chinese authorities remove over 47,000 ai-generated celebrity content pieces in crackdown

Cyberspace authorities have reported the removal of over 47,000 pieces of non-compliant and illegal content, alongside the handling of 3,961 accounts, as part of a recent enforcement action against AI impersonation of public figures. The prohibited material primarily involved monetization schemes, such as unauthorized use of AI to mimic celebrities' likenesses or voices, fabricating rumors by altering portraits, or generating explicit content for financial gain. These violations spanned major platforms including WeChat, Douyin, Xiaohongshu, Bilibili, and Kuaishou. This operation is part of an ongoing governance effort by cyberspace authorities concerning AI-related disorder in 2026. Earlier this year, in September, the Cyberspace Administration of China initiated the second phase of its Qinglang campaign, which focused heavily on imitation, having already removed over 5.61 million pieces of content and managed more than 49,000 accounts by early September. Furthermore, at the close of September, a nationwide two-month Qinglang campaign targeting fake personas in short videos was launched, specifically prioritizing the crackdown on fake endorsements and marketing utilizing synthesized audio and video.