Legal peculiarities of watermarking of content created by generative AI in PRC

This work is dedicated to analyzing the problem of labeling materials created by artificial intelligence, as well as studying approaches to this issue in the People’s Republic of China (PRC). The study provides data on the features of generative artificial intelligence’s work with large datasets. The problem of information degradation, the accumulation of errors within it, and the consequences of the uncontrolled spread of synthesized content on the internet are examined. The researchers also explore the issue of the “knowledge collapse.” The main approaches of Chinese legislators regarding the regulation of this issue are outlined. Several specific technical mechanisms, contained in the regulatory documents of the PRC, related to embedding hidden and explicit labeling in AI-generated content, are noted.