China Targets AI, Wants More Control

On Saturday, the Cyberspace Administration of China announced a draft plan that would the obligation to mark content generated using AI. This is to apply to static texts, graphics, videos, and audio tracks. All of them are to have a layer clear watermark.

The entry into force of this law is almost certain

Currently, it is only a “preliminary plan”, but the first actions will reportedly be taken in October. This is to give time for the largest platforms such as BiliBili (the Chinese equivalent of YouTube). The obligation will also apply to private individuals.

Of course The Chinese Communist Party Doesn’t Care About Artists. These are simply further actions aimed at increasing censorship and surveillance Chinese Internet. This is indicated, for example, by the proposal to provide personal data when using content generators, which will be stored for a period of 6 months. This is to allow eliminating any form of criticism of the government with the help of AI.

Some AI-created content labels will be applied dynamically, based on metadata in files. In the absence of these, service providers will be required to analyze the content. If a company suspects that a file was created by AI, it will have to mark it.

Theoretically, the project proposal is open, and Chinese netizens have the opportunity to express their opinions on it. However, it is no secret that the Chinese Communist Party does not take the citizens’ opinions very seriously. Thus, the entry into force of this law is practically certainand any changes will be minor.