With the rapid development of artificial intelligence, new types of illegal and criminal activities have emerged, posing threats to cyberspace security and safety of people's personal and property interests.

On Oct 14, Oriental Horizon, a TV program produced by China Central Television's Channel 13, reported on typical AI-related infringement cases adjudicated by the Beijing Internet Court (BIC). BIC judges Sun Mingxi, Yan Jun and Liu Chengzu were invited to the program to analyze such cases.

Using another person's image and sound for commercial promotion

The plaintiff Li is a well-known university professor famous in the field of education and children's development. In 2024, she discovered multiple videos in an online shop selling educational books, in which she appeared to be advertising a series of educational books marketed by a cultural media enterprise.

Li never recorded such promotional videos. She believed that the enterprise had, without authorization, used her personal image, professional background and social influence to attract attention and increase sales, thus infringing on her rights to portrait and voice.

Li sued the enterprise and requested the books sellers bear tort liability for publishing the videos, including making an apology and paying compensation. The defendant admitted that it has used Li's recordings and AI-synthesized voice for promotional purposes, but argued that the videos did not damage the plaintiff's reputation and therefore caused her no loss.

After hearing the case, the court found that the public would easily associate the book promotions in these videos with the plaintiff given the plaintiff's well-known status in education and children rearing, and the sound was similar with her timbre, intonation and pronunciation. Therefore, the publication of the videos constituted an infringement of Li's rights of portrait and voice Evidence presented in the case proved that the defendant had failed to fulfill its duty of reasonable review and should bear joint liability with the video publishers for the infringement in the case. The court ruled that the defendant should issue an apology to the plaintiff and compensate the plaintiff 120,000 yuan ($17,040) for her economic loss and reasonable expenses incurred for protecting her rights. The judgment has taken effect.

Creating celebrity's AI companions: What rights are infringed?

An AI technology enterprise developed and operated a mobile accounting application, allowing users to create "AI companions" by setting their names, avatars, and the type of relationship with users, such as romantic partners or family members.

The plaintiff He is a famous television host and actor, and a widely recognized public figure. In this app, many users created AI companions using He's name and portrait, uploaded his photos as avatars, and set various relationship identities. Through algorithm deployment, the defendant categorized these AI companions based on some relationship settings and recommended the AI character He to other users. He filed a lawsuit against the enterprise, requesting a public apology and compensate him for his economic loss and emotional distress.

After hearing the case, the BIC found that, under the defendant's app functions and algorithm design, users used the plaintiff's name and portrait to create the AI character produced interactive materials, and projected an overall image integrating his name, portrait and personality traits onto the AI characters. This constituted the use of He's overall personality image containing his name and portrait, without his authorization, and therefore infringed upon his rights to name and portrait. In addition, users were can freely set their virtual relationships with He's AI character and use word or sound reaction materials to "train" the character, resulting in a high degree of association between the AI characters and the real natural person. Such use, without He's consent, infringed upon his personal dignity and personality freedom, constituting an infringement of his general personality rights. The court ruled that the defendant should issue a public apology and compensate He 203,000 yuan for his emotional distress and economic loss.

The judge explained that a natural person's personality rights extend to their virtual image. Creating and using a virtual image of a natural person without authorization constitutes an infringement to a natural person's personality rights. Internet service providers, if participating in the infringement through algorithm design, should bear corresponding tort liability.

Using AI toots to deface another person's image

The plaintiff Cheng and the defendant were members of the same WeChat group. Without Cheng's consent, the defendant used AI software to generate sexually explicit anime-style images based on Cheng's normal profile photo used on WeChat, and posted them in the group. The defendant continued this behavior despite the plaintiff's repeated objection. He produced another image depicting a character with exposed body parts and physical deformities, and sent it to Cheng via private message.

The plaintiff believed that the images could be recognized as herself and bore serious sexual implications and stigmatization, diminishing her social standing and infringing upon her rights of portrait, reputation and general personality rights. Therefore, she requested the defendant to issue an apology and compensate for economic losses and one yuan for emotional distress.

After hearing the case, the court found that the infringing images posted in the WeChat group were generated by the defendant using AI technology based on Cheng's profile photo. Group members could identify the images as depicting Cheng based on physical features and the context of the group chat. Therefore, this defendant's behavior infringed upon the plaintiff's right of portrait. The defendant used AI software to create sexually explicit images based on the plaintiff's normal image, triggering uncomfortable discussions and vulgar comments. Therefore, it infringed upon the plaintiff's portrait and reputation rights. Sending the sexually explicit and deformed image to the plaintiff in a private message also infringed on her general personality rights.

The court ruled that the defendant should post a public apology in his WeChat Moments for 48 consecutive hours and issue a written apology to the plaintiff. The defendant should also compensate her one yuan for her emotional distress, according to the court. The judgement has now taken effect.

Face-Swapping in videos

The Plaintiff Liao is a popular Chinese-style short-video blogger and Hanfu model who has a large number of fans on Douyin. She discovered that a face-swapping and costume-changing application operated by a technology and culture company provided videos identical to her own makeup and attire as templates for users to perform face-swapping. The app advertised that users could complete face-swapping with just one photo through automatic facial recognition. Liao believed that the enterprise's conduct infringed on her rights of portrait and personal information. She filed a lawsuit against the enterprise and requested a written apology and compensate her for economic losses and emotional distress.

The defendant argued that the video had lawful sources and that the facial features in the swapped videos did not identify Liao, and therefore did not infringe upon her right of portrait. The defendant also said claimed that the face-swapping technology was provided by a third party, so it didn't infringe on her right of personal information.

After hearing the case, the court found that although the defendant used the plaintiff's video as a template, it did not constitute an infringement of the plaintiff's right of portrait as the public would identify the replaced faces as third parties rather than Liao, and the use did not involve defamation, distortion or fabrication of her portrait. However, the defendant collected the video that contained the plaintiff's facial information and replaced it with others' faces. This process required the integration of facial traits and expressions between users' pictures and original video through algorithms, thus unavoidably collecting using and analyzing the plaintiff's personal information without her consent. Therefore, the defendant infringed on the plaintiff's right to her personal information.

Sun explained that the portrait right protects personal characteristics that can be recognized by others and that carry individual dignity, while the right of personal information focuses on protecting people's rights and interests in digital space. Where personal information is processed in digital space without authorization or other statutory grounds, such conduct constitutes infringement.

The court noted that the plaintiff is only the model in the videos concerned, and not the video rights holder. Therefore, any infringement claim of creative works should be brought separately by the rights holder. The court ruled that the defendant should issue a written apology to the plaintiff and compensate her for emotional distress. The ruling has now taken effect.