news

한어Русский языкEnglishFrançaisIndonesianSanskrit日本語DeutschPortuguêsΕλληνικάespañolItalianoSuomalainenLatina

with the rapid development of artificial intelligence, ai voice, ai image, and ai video have begun to emerge, and corresponding legal issues have also arisen one after another. zhao ruigang, member of the party leadership group and vice president of beijing internet court, and sun mingxi, a full-time member of the trial committee, were guests on cctv1's "today's statement" to interpret the country's first ai-generated voice personality rights infringement case and ai companionship case.

the first case in china

ai-generated voice personality rights infringement case

the common saying "you can hear the voice before you see the person" shows that everyone's voice has a certain degree of recognition. with the rapid development of artificial intelligence technology, artificial intelligence can also edit and create a person's voice.

ms. yin is a dubbing artist. around may 2023, she discovered that many videos using her ai voice for dubbing appeared on the internet platform. the ai ​​voice involved in the case is a product of a certain workshop. users input the text they want to dub, and the product can convert the text into the corresponding sound. ms. yin once worked with a cultural media company to record two audiobooks, and the company provided the recordings to a software company. a software company used ms. yin's recordings as materials for ai processing, generated a text-to-speech product, and sold it through distributors. in the end, a technology company where the dubbing software certain workshop was located purchased this product, generated a text-to-speech product without technical processing, and sold it on the software. ms. yin believed that her voice rights were infringed, and she sued five companies, including a cultural media company that transferred the recordings to others and a technology company that operated the certain workshop software, to the beijing internet court, demanding that the infringement be stopped, an apology be made, and compensation for economic losses of 600,000 yuan.

the voice of a natural person is distinguished by voiceprint, timbre and frequency, and has the characteristics of uniqueness, uniqueness and stability. after the in-court inspection, the judge believed that the voice of the ai ​​and ms. yin's timbre, intonation, pronunciation style, etc. were highly consistent, and the ai ​​voice could be associated with ms. yin's voice. the ai ​​voice involved in the case is recognizable and should be protected by law.

as a personal right, voice is personal and exclusive. having the copyright to the recordings made by ms. yin does not automatically mean owning the rights to ms. yin's voice. the defendant, a cultural media company, processed and used her voice through ai without ms. yin's authorization, which constitutes infringement.

a technology company that operates a certain workshop product said that it purchased the text-to-speech product involved in the case from a distributor. it was just a normal purchase and paid a reasonable price. it did not have excessive audit obligations. the court held that it only had the responsibility to stop the infringement. after ms. yin filed a lawsuit, a certain technology company removed ms. yin's ai voice from the shelves, and the infringement has stopped.

in this infringement chain, a cultural media company that provides audiobooks and a software company that ai-processed ms. yin's voice should be responsible for the infringement and bear the corresponding compensation liability. in the end, the court comprehensively considered the circumstances of the infringement, the value of similar market products, the volume of product playback and other factors, and ordered a cultural media company and a technology company to compensate the plaintiff ms. yin for a total of 250,000 yuan. after the first-instance verdict, neither party appealed. this case is the first case of infringement of personality rights of ai-generated voices in the country.

the development of science and technology will bring about some legal issues. the judges hope that this case will trigger everyone's thinking. as a natural person, you need to pay attention to protecting your voice; for companies involved in related industries, they should pay special attention to obtaining authorization, otherwise it is easy to trigger a series of lawsuits. nowadays, some audiobooks and audio-visual products have begun to use ai voice dubbing, which reduces costs and brings convenience. however, at the same time, it is necessary to abide by the legal bottom line.

ai companionship case

if you spend 80 yuan on a movie, your "idol" will ask you who you went to the movie with, and even look a little jealous; if you are sick, your "idol" will say that he feels sorry for you and ask you to take care of yourself. this is a mobile accounting software, and its special feature is that users can create, add, and chat with ai companions. this companion can be a virtual character or a person in real life. the software will also automatically push some well-known public figures to attract users.

he is a public figure who found that he had been set up as a virtual companion by many users. after selecting an ai companion in this accounting software, users can also freely set the relationship and title between each other. after the user records the account in the software, the ai ​​companion will send corresponding comments, emoticons, etc., respond in an anthropomorphic way, and even ask about the user's well-being. after learning about some of the settings of this software, he felt that the software was suspected of infringement. the accounting software is operated by an artificial intelligence technology company in shanghai. he's lawyer subsequently sent a letter to the company to communicate, but the company believed that they had not infringed and was not responsible. he filed a lawsuit with the beijing internet court, demanding an apology and compensation for economic losses.

the defendant, a shanghai artificial intelligence technology company, said that it should not be held responsible because he's pictures, emoticons and even personalized reply texts were uploaded and set by other users, and the defendant was only a service provider. after receiving he's lawyer's letter, the defendant promptly deleted he's related content in the software and should no longer be held liable for infringement.

the defendant said that if multiple users apply for the same ai companion, the system will open the "training" function. users enter the training page to enter the desired replies, emoticons, etc., and the software will record and save them. for example, a user saved he's image emoticon package and trained the character. the system entered the content into the relevant database. when the next user records the account, the system will send customized replies and emoticons after receiving certain keywords. if an ai companion has been trained more, the language he uses to reply to users may be richer. the judge said that he could see that the replies of the ai ​​companion involved in the case were closer to he's real personality.

because the software involves some technical issues, the court invited a special technical investigator to participate in the trial. the technical investigator asked the defendant in detail about the algorithm and push function. after understanding the relevant technical issues, the judge believed that the defendant did not simply provide technical services. the product rule design and algorithm application made by the defendant directly determined the service core and functional realization of this product. the software is more like providing content rather than a simple channel. therefore, the defendant, as a content service provider, should bear the infringement liability.

the defendant used he's name and portrait without he's permission, which constituted an infringement of he's right to name and portrait. in addition, the ai ​​companion also systematically used he's overall image, which also violated he's general personality rights. finally, in august 2021, the beijing internet court ruled that the defendant, a shanghai artificial intelligence technology company, should publicly apologize to the plaintiff he and compensate for economic losses of 183,000 yuan and mental losses of 20,000 yuan. this case was selected by the supreme people's court as a "typical civil case of judicial protection of personality rights after the promulgation of the civil code."

the judge pointed out that although the virtual images on the internet are not real entities, publishing relevant remarks and using their images may involve the derogation of other people's personal dignity. in particular, the overall image of some public figures has great commercial value, and unauthorized use without permission is likely to be suspected of infringement. enterprises should always pay attention to the boundaries of the law during their innovation and development. while internet users use new technologies to meet their personal emotional and cultural needs, they must also fully respect the personal rights and interests of others.

editors: ren huiying, liu wanyue