China to require labels for AI-generated content as tech brings fresh challenges
Beijing proposes rules to fight misinformation, deepfakes, copyright issues and other risks from media made with artificial intelligence
Online content providers would be required to use conspicuous labels – such as text, sound or graphics – for all AI-generated text, video, audio and virtual scenes.
Meanwhile, more subtle identifiers, such as digital watermarks and metadata tags, would also be encouraged for AI-generated content.
The draft also stipulates that no organisation or individual may maliciously delete, tamper with, forge, or conceal the required labels or violate the rights and interests of others through improper identification of AI-generated content.
The proposed rules further require online content platforms to regulate any AI-generated content they distribute.
According to watchdog body the Cyberspace Administration of China, the guidelines will apply to industry organisations, enterprises, educational and scientific research institutions, public cultural institutions and professional institutions that use AI to generate content and provide services to the Chinese public.