China’s Social Media Platforms Begin Labelling AI Content

China has introduced a law that makes it mandatory for social media platforms to add clear labels to material created with artificial intelligence. The rule took effect on Monday and applies to popular services including WeChat, Douyin, Weibo and Xiaohongshu. Users will now see labels on posts containing text, images, audio, video or other content generated by AI.

Visible and hidden identifiers

The regulation requires two types of markings. Some must be visible to viewers on the post itself. Others must be hidden, such as digital watermarks stored in the metadata. The law also bans any attempt to remove or change these labels. In addition, platforms must block AI material that spreads illegal activity, violates rights or misleads the public.

Platform responses

WeChat has told its users they are responsible for declaring AI content before posting. The platform also said it would stop people from deleting or tampering with labels that it applies automatically. Douyin has encouraged creators to place visible tags on their posts, while also using metadata checks to confirm the origin of content. Weibo has added a reporting tool so that users can flag posts that should have labels but do not. Xiaohongshu reminded its community that it will add its own identifiers if users fail to do so.

Regulatory background

The law was drafted earlier this year by the Cyberspace Administration of China together with three other ministries, covering industry, public security and broadcasting. Officials said the measure is part of wider efforts to manage risks linked to AI content. These include deepfake technology, which can alter images, speech or video in ways that affect both privacy and security.

Part of a broader campaign

The new labelling rule connects with a broader campaign to strengthen control over online spaces. Earlier this year, regulators launched a review of AI services, aimed at limiting misinformation, copyright abuse and misleading commercial practices. This has been tied to the “clear and bright” initiative, an annual programme that seeks to reduce harmful or deceptive activity online and provide more safeguards for young users.

Global comparisons

China is among the first countries to make AI labelling a legal requirement. In other regions, some technology companies have taken voluntary steps. Certain new smartphones now include tools that can attach digital credentials to images and videos, allowing people to confirm how they were made.

Notes: This post was edited/created using GenAI tools. Image: DIW-Aigen.

Read next:

• Global Startup Cities in 2025: Power Shifts Beyond Silicon Valley^[2]

• Daily Vitamin D May Keep Cells Healthier for Longer, but Scientists Urge Caution^[3]

^[1]