China Drafts Groundbreaking Rules to Regulate AI Digital Humans and Protect Minors

China's top internet regulator has taken a pioneering step in AI governance by publishing draft rules that specifically target "digital humans" — AI-driven virtual personas capable of holding conversations, forming ongoing relationships, and generating personalized content. The Cyberspace Administration of China (CAC) released the proposed framework on April 3, 2026, marking one of the first attempts globally to regulate the interaction layer between humans and AI rather than just the models themselves.

Key Highlights

• Mandatory labeling: All digital human content must carry a prominent disclosure so users can distinguish AI personas from real people
• Minors protection: Virtual intimate relationships with users under 18 are explicitly prohibited
• Mental health monitoring: Platforms must identify and assist users showing suicidal or self-harm indicators
• Consent requirements: Digital humans cannot be created using a person's likeness or data without explicit consent
• Identity verification: Virtual personas are barred from bypassing identity verification systems

What Are Digital Humans?

Digital humans are AI-powered virtual personas that go beyond simple chatbots. They can maintain persistent identities, remember past conversations, develop relationship dynamics with users, and appear as realistic avatars in live streams, customer service, education, and entertainment. In China, these virtual personas have proliferated across e-commerce livestreaming, companionship apps, and educational platforms.

Why This Matters

Unlike previous AI regulations that focused on model training data or content generation, these draft rules target the emotional and psychological relationship between AI systems and their users. The CAC is essentially drawing boundaries around how AI can interact with humans on a personal level — a regulatory frontier no other country has formally addressed.

The mental health monitoring requirement stands out as particularly ambitious. Platforms deploying digital humans would be required to detect signs of emotional distress or self-harm among users and intervene accordingly. This goes beyond what current U.S. federal regulations require from AI platforms.

Impact on the Industry

The proposed rules carry significant implications for companies operating AI companion services. Platforms like Baidu and Tencent, which have invested heavily in digital human technology for livestreaming and customer engagement, will need to implement new compliance layers.

For international companies, the regulations signal a tightening environment for anyone deploying AI companion or avatar services targeting Chinese users. The consent and identity verification requirements also directly address growing concerns around deepfake technology.

Content and Security Restrictions

Beyond user protection, the draft includes familiar content controls: digital humans are prohibited from generating material that endangers national security, promotes secession, incites ethnic discrimination, or distributes sexually suggestive or violent content. Service providers face a dual compliance burden — child safety requirements alongside ideological content screening.

What Comes Next

The public comment period remains open until May 6, 2026. The CAC has not announced a finalization timeline, and no specific penalties for violations have been detailed in the current draft. However, given China's track record of moving swiftly from draft to enforcement on AI regulations, companies are expected to begin compliance preparations immediately.

This regulatory move could set a precedent for other countries wrestling with how to govern increasingly human-like AI systems. As digital humans become more realistic and emotionally engaging, the question of where to draw the line between helpful AI and manipulative technology is one every government will eventually need to answer.

Source: US News & World Report