Author

Ruan Shenyu, Assistant Professor, Renmin University of China Law School; Research Fellow, Law and Technology Institute; Doctor of Juridical Science (J.S.D.)

Abstract

The requirement of portrait identifiability is crucial in disputes involving images with "removed facial features," and its significance has increased with the rise of AI-generated virtual avatars.Existing "holistic assessment” approaches rely heavily on judicial discretion,creating uncertainty and overlooking the need for differentiated identifiability standards that correspond to the distinct legal interests protected by portrait rights.These interests include dignity,identity,and commercial value.The standard of identifiability should therefore vary accordingly:for dignity interests,what should apply is the most permissive standard—identifiability exists if the rights holder himself can recognize the image;for identity interests,the test is whether people familiar with the rights holder can confidently conclude that the image refers to that person alone;and for commercial interests,the strictest standard applies,requiring that the general public,under isolated comparison,can immediately recognize the person depicted.This framework supports a more nuanced and coherent approach to portrait identifiability.

Keywords: portrait rights ; artificial intelligence ; identifiability ; typology