The Cyberspace Administration of China has rolled out new labelling measures for AI-generated content, which will take effect on September 1, 2025. This article dives into the details of these measures and compares them to the upcoming EU AI Act.
Key Points
- New labelling measures include explicit and implicit labels for AI-generated content.
- Explicit labels must be included where there’s a risk of public confusion, such as in text or deep fakes.
- Implicit labels involve metadata for all AI-generated content, with encouragement to use digital watermarks.
- Internet propagation service providers face new obligations for regulating the distribution of AI content.
- Differences between China’s rules and the EU AI Act reflect varying degrees of compliance burdens in the AI landscape.
Content Summary
On March 14, 2025, the Cyberspace Administration of China published the Labeling Measures for AI-generated content, setting standards that public entities must follow starting September 1, 2025. These rules require AI content providers to label their synthetic content clearly, either through explicit labels when public confusion is a risk or implicit labels in file metadata for all AI-generated products.
Providers include AI content generation services, internet content propagation services, and app distribution platforms, all of which share specific responsibilities for compliance. Users publishing AI-generated content are also mandated to declare its nature actively.
In contrast, the EU AI Act, effective August 2, 2026, shows a more lenient structure, placing labelling obligations primarily on AI system providers rather than on third parties. The differences highlight the complexity organisations face when navigating the regulatory environments in various jurisdictions.
Context and Relevance
The new AI content labelling rules in China are particularly significant given the increasing global focus on AI transparency and ethical practices. As businesses increasingly rely on AI for content creation, understanding these regulations becomes crucial for compliance and risk management. The contrasting frameworks in China and the EU reveal the potential challenges for international organisations operating in both regions, making familiarity with these laws a key asset in today’s AI-driven market.
Why should I read this?
If you’re in the content creation game or just curious about how different countries are handling AI regulations, this piece gives you the lowdown on what’s brewing in China compared to the EU. It’s a big deal because these rules shape how AI-generated content is labelled and perceived, and you don’t want to be caught off guard by what lies ahead!
