China is doubling down on regulating AI-generated content with a new mandate set to take effect on September 1, 2025. The Cyberspace Administration of China (CAC) now requires that all AI-generated or synthetic content—including text, images, videos, and audio—must have explicit labels both visibly and in metadata. The move is aimed at preventing misinformation and ensuring users can distinguish between human-created and AI-produced media. Notably, app stores will also need to verify compliance, and tampering with or removing these AI labels will be strictly prohibited. While China is known for its tight control over the internet, this law aligns with global efforts like the EU’s Artificial Intelligence Act to maintain transparency in AI use. However, how strictly it will be enforced—and what penalties violators could face—remains uncertain.

Key Points:

• Clear AI Labeling: All AI-generated content must have visible or audible labels and embedded metadata.
• Types of Affected Content: Includes text, images, videos, audio, and virtual environments.
• User Responsibility: If users request unlabeled AI-generated content for industrial or social needs, this must be logged, and it becomes their responsibility to ensure compliance.
• Tampering is Illegal: Removing, falsifying, or hiding AI labels is prohibited, as is falsely labeling human-made content as AI-generated.
• App Store Accountability: Apps that generate AI content must ensure compliance before being listed.
• Combatting Misinformation: China sees this move as essential to preventing AI-driven disinformation.
• Comparison to EU AI Act: Similar regulations exist in Europe as part of broader AI governance measures.
• Enforcement Uncertainty: No prescribed punishments have been outlined, but legal action remains a possibility.

China’s AI labeling rule may be stringent, but it reflects a growing global consensus that AI-driven content needs transparency. Whether other nations follow suit—or how AI creators respond—will be interesting to watch.
Link to Article