China is set to require clear labeling of all content generated by artificial intelligence (AI) under new regulations set to take effect later this year.
The directive, issued by the Cyberspace Administration of China (CAC) along with several other government ministries, aims to combat the spread of misinformation and increase transparency on the internet as the nation rapidly adopts the new technology. The rules will go into effect on September 1.
The new rules stipulate that all AI-generated text, images, audio, video, and virtual content must be clearly labeled, both explicitly and implicitly. Explicit labels should be easily visible to users, and implicit identifiers, such as digital watermarks, must be embedded within the metadata.
Online service providers involved in the creation of AI content are also mandated to comply with China’s cybersecurity and deep synthesis management rules. Platforms are required to verify content made with AI before publication and to add labels as needed.
If metadata lacks AI markers, but the content shows signs of AI generation, platforms must flag this accordingly. Additionally, app distribution platforms will be tasked with assessing AI-related features before approving services.
