Chinese social platforms rush to abide by AI-generated content labelling law (www.scmp.com)

China’s social media platforms are rapidly implementing a new law requiring clear labelling of AI-generated content, including text, images, audio, and video. Enacted in March, the regulation mandates both explicit labels that are visibly displayed to users and implicit identifiers—such as digital watermarks embedded in metadata—to ensure transparency about AI involvement. This move underscores Beijing’s heightened focus on controlling AI’s impact amid growing concerns around misinformation, copyright violations, and online fraud. The law is part of the Cyberspace Administration of China’s (CAC) broader 2025 Qinglang campaign, which aims to “clean up” the digital environment by strengthening oversight over emerging technologies. Leading platforms like WeChat (Weixin) have announced that content creators must voluntarily disclose AI-generated material, with the platform also prompting users to critically assess unlabelled content. For the AI/ML community, this signals increased regulatory scrutiny and a push for accountability in AI content creation—paving the way for broader adoption of technical measures like watermarking to authenticate AI outputs and combat misuse.