01
Factual background & dispute
- Users could create AI companions and define names, portraits, and relationships.
- Many users created companions based on the claimant and uploaded the claimant's portrait and conversational materials.
- The operator clustered and recommended the roles and organized user-submitted material into a dedicated dialogue corpus.
02
Core issues & judicial focus
- How a virtual persona combining a person's name, portrait, and personality traits is protected
- How responsibility is allocated between user creation and platform organization and recommendation
- Whether simulated companionship with a real person requires that person's consent
03
Judicial finding & holding
- A natural person's virtual persona incorporates protected elements including name, portrait, and personality traits.
- The platform's classification, recommendation, and organization of dialogue materials went beyond passive technical hosting.
- The operator was ordered to apologize and pay RMB 203,000 for economic loss, reasonable expenses, and emotional harm; the first-instance judgment became effective after withdrawal of the appeal.
04
Practical risk implications
01Companion products require express authorization before using a real person's name, image, voice, or persona.
02User creation does not eliminate platform responsibility for clustering, recommendation, training, and commercialization.
03Products should support identity complaints, character removal, data deletion, and repeat-creation prevention.