Draft Administrative Measures for Digital Virtual Human Information Services

Analysis Of China’s Draft Administrative Measures For Digital Virtual Human Information Services 

On April 3, 2026, the Cyberspace Administration of China released the Administrative Measures for Digital Virtual Human Information Services (Draft for Comments), a pioneering regulatory framework addressing AI-enabled infringements of personality rights such as name, image, and reputation under the Civil Code. The Draft Measures defines digital virtual humans broadly to encompass both motion-capture-driven and computationally generated entities, and establishes a tripartite classification of regulated actors: Digital Virtual Human Service Providers, Technical Supporters, and Service Users. Its core provisions, concentrated in chapters on Rights Protection and Service Standards, impose obligations covering general personality rights, personal information protection aligned with the PIPL, intellectual property, minor protection and anti-addiction measures, risk monitoring, technical and human oversight, enforcement actions including service termination and virtual human deregistration, user grievance channels, mandatory service agreements across the value chain, algorithmic filing for entities with public opinion or social mobilization attributes, and a tiered penalty regime reaching fines of up to 200,000 yuan. Situated alongside existing instruments on algorithmic recommendation, deep synthesis, generative AI services, and AI-content labeling, the Draft Measures reflects China's proactive, sector-specific governance approach, though the author argues it would benefit from proportionality-based differentiation by entity size and risk modeled on the EU Digital Services Act, clearer harmonization with Civil Code notice-and-takedown mechanisms, graduated implementation with review cycles, and a more granular classification system that better matches obligations to the distinct roles and capabilities of providers, technical supporters, and users. 

1. Regulatory trigger: AI technology has increasingly been exploited to infringe personality rights under the Civil Code, prompting the Cyberspace Administration of China to release the Draft Measures on April 3, 2026.

2. Broad definition: a digital virtual human is any non-physical virtual image built with computer graphics, digital image processing, or AI that simulates human appearance, voice, behavior, interaction, or personality.

3. Tripartite classification: the Draft Measures regulates Service Providers (DSPs), Technical Supporters (DTSs), and Service Users (DSUs), allocating obligations along the value chain.

4. Rights protection framework: Chapter II safeguards general personality rights, personal information under the PIPL, intellectual property, and minors through anti-addiction measures.

5. Service standards: Chapter III requires risk monitoring, emergency response, content management systems, and combined AI and human review scaled to the provider's size.

6. Enforcement duties: DSPs must apply dynamic identity verification, warnings, feature restrictions, service termination, and, for serious risks, suspension, deregistration, and elimination of effects.

7. Grievance and contracting: DSPs must offer user complaint channels with timely feedback and enter into service agreements specifying content security and data-handling standards with DTSs and DSUs.

8. Algorithmic accountability: entities with public opinion or social mobilization attributes must comply with algorithmic filing and change procedures under China's algorithmic recommendation rules.

9. Tiered penalties: Article 24 escalates from warnings and rectification orders to fines of 10,000 to 100,000 yuan, and up to 200,000 yuan where life or health is endangered.

10. Regulatory ecosystem: the Draft Measures complements existing rules on algorithmic recommendation, deep synthesis, generative AI services, and AI-content labeling.

11. Proportionality concern: a one-size-fits-all approach risks overburdening smaller platforms, and a tiered model similar to the EU Digital Services Act is recommended.

12. Harmonization gap: the relationship between emergency response duties and the Civil Code's notice-and-takedown mechanism needs clarification to avoid expanded platform liability.

13. Graduated implementation: staged rollout with review mechanisms would allow the framework to adapt to rapid technological change.

14. Granular responsibility allocation: future rules should match obligations more precisely to the distinct roles, capabilities, and risk profiles of DSPs, DTSs, and DSUs.