AI governance in China.
China has built the world's most comprehensive suite of AI-specific regulations — governing recommendation algorithms, deepfakes, generative AI, and AI labelling in separate instruments. CAC enforcement is active and consequential. Chinese content available at /zh.
China's AI regulatory stack
Five regulations, all active, all enforced by CAC.
Applies to algorithm recommendation service providers. Disclosure of recommendation algorithms required. Users can opt out of personalised recommendations.
AI-generated synthetic audio/video/image/text must be labelled. Real-name registration for high-risk deep synthesis services. Providers must file with CAC.
Publicly available generative AI services in China must file with CAC. Training data must be lawful and accurate. Content must not violate core socialist values, endanger national security, or be false information.
Explicit labels (visible to users) and implicit labels (machine-readable metadata) required for AI-generated content across text, audio, images, and video.
Comprehensive data protection. Consent-based. Cross-border transfer restrictions require CAC security assessment for large-scale processors. Automated decision-making transparency and contestation rights.
Stay ahead of AI governance
Regulatory updates, practical frameworks, and analysis. No spam, unsubscribe anytime.
No spam. Unsubscribe anytime. We'll never share your email.