China’s new regulations on AI-generated content labeling
These regulations mandate that internet service providers and online platforms clearly identify AI-generated content to enhance transparency and prevent misinformation.
On 14 March 2025, the Cyberspace Administration of China (CAC) announced the ‘Measures for Labeling Artificial Intelligence-Generated Content,’ set to take effect on 1 September 2025. These regulations mandate that internet service providers and online platforms clearly identify AI-generated content to enhance transparency and prevent misinformation.
The guidelines introduce two primary labelling methods:
- Explicit labels: Visible markers (e.g. text or symbols) indicating content is AI-generated.
- Implicit labels: Metadata embedded within the content, detailing its AI origin.
Online platforms are also required to develop mechanisms to detect and label AI-generated content, categorising it as confirmed, possible, or suspected AI-generated material.
Understanding AI labelling
AI labelling refers to the process of identifying and marking content that has been generated or synthesised using AI technologies. This practice is crucial for maintaining transparency, ensuring users are aware of the origins of the content they consume, and preventing the spread of misinformation. By implementing both explicit and implicit labels, stakeholders can uphold the integrity of information and foster trust in digital platforms.
