TECH

China Enforces AI Content Labeling Rules to Curb Misuse

Published: Sep. 3, 2025 4:59 a.m. GMT+8

00:00

00:00/00:00

Listen to this article 1x

ByteDance Ltd.'s Douyin application on a smartphone. Photo: Bloomberg

China’s long-anticipated rules requiring artificial intelligence (AI)-generated content to be labeled took effect Monday, forcing tech companies and social media platforms to clearly mark material made with AI.

The new regulation, jointly issued by the Cyberspace Administration of China, the Ministry of Industry and Information Technology, the Ministry of Public Security and the National Radio and Television Administration, mandates both explicit and hidden labels for synthetic content.

You've accessed an article available only to subscribers

Subscribe today for just $.99.

VIEW OPTIONS

Unlock exclusive discounts with a Caixin group subscription — ideal for teams and organizations.

Subscribe to both Caixin Global and The Wall Street Journal — for the price of one.

Share this article

Open WeChat and scan the QR code

DIGEST HUB

Digest Hub

Back

Explore the story in 30 seconds

China’s AI content labeling rules took effect Sept. 1, 2025, requiring both explicit (watermarks, notices) and hidden (metadata) tags for AI-generated media.
Major platforms like Douyin, DeepSeek, Zhipu AI, and SenseTime announced compliance steps, updating tools and policies to detect, label, and trace AI content.
The regulation prohibits tampering with labels and is part of broader efforts to curb AI misuse and protect users from deceptive content.

AI generated, for reference only

Who’s Who

Zhipu AI: Zhipu AI has implemented measures to comply with China's new AI content labeling regulations. Its Qingyan chatbot now displays "AI-generated" notices for responses. When text is copied, a disclaimer is appended. Zhipu AI also uses both explicit watermarks and hidden digital identifiers for all its text, images, and videos.

DeepSeek: **DeepSeek** is a Chinese AI developer that has implemented disclaimers on its mini-app to comply with new regulations. These disclaimers state "content generated by AI, for reference only." DeepSeek also warns users against altering or removing labels and spreading false or illegal information using AI.

SenseTime Group Inc.: SenseTime Group Inc., a prominent AI developer in China, has implemented "AI-generated" tags to watermark its digital humans. Additionally, conversations with its chatbot now include disclaimers. The company has also made APIs and documentation available to assist partners in integrating labeling features.

Douyin: Douyin, ByteDance's short-video app, has implemented new guidelines requiring creators to declare their use of AI. If creators fail to do so, Douyin will use technical tools to detect and label the content with a "suspected AI-generated, please verify carefully" tag. Additionally, Douyin will incorporate hidden labels for end-to-end traceability, building upon its 2023 platform rules demanding AI labeling and real-name registration for virtual humans.

ByteDance Ltd.: ByteDance Ltd., owner of the short-video app Douyin, is implementing measures to comply with China's new AI content labeling rules. Douyin will require creators to declare AI use and will technically detect and label suspected AI-generated material. To enhance traceability, Douyin will also incorporate hidden labels. These guidelines build upon existing platform rules from 2023.

Xiaohongshu: Xiaohongshu, a lifestyle app, has implemented a "three-step guide" for tagging AI-generated posts. They also reserve the right to add labels themselves if content creators do not self-declare AI-generated material. This is part of a broader regulatory push in China to require clear labeling of AI-generated content on platforms.

AI generated, for reference only

What Happened When

2023：: Douyin released 11 platform rules requiring AI labeling and real-name registration for virtual humans.

September 2024：: The measure requiring AI-generated content labeling was first floated as a draft.

March 2025：: The regulation for labeling AI-generated content was formally issued.

April 2025：: Regulators launched a three-month nationwide campaign called 'Clear and Bright' targeting AI misuse.

August 2025：: Douyin issued guidelines instructing creators to declare AI use; Xiaohongshu rolled out a 'three-step guide' for tagging AI-generated posts.