Shanghai’s Cyberspace Administration has directed platforms including Xiaohongshu,Watch Serve the People Online Bilibili and PDD to remove unauthorized AI-generated products and content as part of a national campaign to curb AI misuse, according to an official statement. Since late April, the regulator has focused on six types of violations, such as AI tools that undress people with one click, unauthorized face or voice cloning, and unregistered AI services. More than 820,000 illegal posts have been removed, more than 1,400 accounts penalized, and about 2,700 noncompliant AI agents taken down. Platforms such as Xiaohongshu and Bilibili have also launched reporting channels for harmful AI content. Authorities have emphasized compliance among developers offering generative AI services, with 82 large models and 87 applications registered so far. Three noncompliant services have faced penalties after failing to meet registration requirements. [Shanghai Cyberspace Administration, in Chinese]
