China's Cyberspace Administration announced comprehensive new regulations today requiring all AI-generated content distributed within the country to carry visible watermarks and metadata labels identifying it as machine-produced.
The regulations apply to text, images, audio, and video generated by AI systems and take effect on July 1, 2026. Platforms that fail to enforce labeling face fines of up to 50 million yuan and potential suspension of operating licenses.
The rules represent the most aggressive AI content labeling regime in the world, surpassing similar but less comprehensive requirements being considered by the EU and United States. Chinese AI companies including Baidu, Alibaba, and ByteDance have pledged compliance.
