Judge reminds - Be wary of the legal risks of misuse of AI technology

Artificial intelligence technology has been quietly integrated into every corner of our lives in recent years: common functions such as intelligent voice customer service, AI anchors, and voice-to-text have greatly improved the convenience of work and life. Many AI face-changing, AI painting apps or small programs are also easy to use, allowing ordinary people to realize their dreams of becoming a star or a painter in one second. ChatGPT, a chat tool developed in the United States at the end of last year, became very popular around the world once it was released. However, technology is always a double-edged sword. Following the "AI Stefanie Sun" cover incident, recently, a man in Fujian was defrauded of 4.3 million yuan by a scammer using AI face-changing and onomatopoeia technology to pretend to be his friend and defrauded him of 4.3 million yuan through a 10-minute video chat. , making people think more deeply: How should AI technology serve mankind? What legal risks will technology misuse bring, and what kind of trust crisis will it cause?

1. AI “creation” has the risk of infringement

The original intention of the development of AI technology is to improve human life. In recent years, its processing of videos, audios, images, and text has gradually moved from the laboratory to the public, changing people's production and lifestyle. It has a wide range of application scenarios and diverse gameplay. It has extraordinary performance in reducing costs, increasing efficiency, and liberating productivity.

Take AI video technology as an example. In the past, a certain actor's misconduct may have caused the entire work in which he participated to be unable to be broadcast, causing financial losses to the filming party. Through AI technology, the actor in question can be "face-changed" after the filming of the work is completed, and the loss can be stopped in time. In terms of special effects production, in the past, AI was only an assistant to special effects artists, but now there are multiple softwares that can generate blockbuster special effects in a "fool-like" way. Movie special effects have entered the "civilian era", and ordinary people can also have the wonderful experience of making their own special effects. In addition, AI technology can also create videos "out of nothing" based on only copywriting without adding other pictures and video materials, meeting the more diversified needs of creators.

During the Spring Festival this year, a set of pictures of fluffy cartoon rabbits with red backgrounds became popular in the circle of friends and became many people’s screensaver wallpapers and avatars. These pictures were all generated by AI drawing technology. Many AI drawing tools have emerged in the fields of design and painting. Users only need to enter text commands to achieve "drawings from text" and get a good work in a few minutes, which greatly improves production efficiency. However, this may have an impact on the image copyright industry.

Generally speaking, if you want to use a copyrighted image, including paintings and photography, you need to pay to use it legally. Otherwise, copyright infringement cases will occur, especially in the field of self-media related disputes. Some self-media accounts have used other people's copyrighted images as illustrations in advertisements without permission, and have been sued by the rights holders for compensation. Some copyright agencies may also deliberately conceal and induce others to use copyrighted images, etc. to obtain the facts of infringement. Later, they ask for compensation from the other party, force them to sign an annual contract or file a lawsuit, turning rights protection into a means of profit-making. The emergence of AI drawing technology can, to a certain extent, help self-media operators control costs and reduce the risk of lawsuits caused by improper use of picture works. On the other hand, it can also disrupt the market order of the commercial rights protection industry chain of picture works. Behavior constitutes a blow.

The above-mentioned AI video, AI drawing technology, and the "all-purpose" chat tool ChatGPT that can help people write papers, emails, scripts, poems, and codes are all specific applications of deep synthesis technology, regardless of whether people have professional backgrounds and related Experience, as long as you have simple expression skills, you can "create" with the help of these AI tools. So, what is the copyright ownership of these generated videos, pictures, and text works? In fact, my country’s copyright law does not clearly stipulate the copyright ownership of works generated using AI. However, many AI tools have set corresponding usage rules for the ownership of the rights of the generated works. Some software also has rules for free users and paid users. The ownership of the rights of the work and whether it can be commercialized are distinguished, which can be regarded as the two parties reaching an agreement on the copyright issues of the generated work, and the user's disposal of the work should comply with the agreement.

In addition, some people also question whether the works generated by AI through learning and imitating massive samples will infringe on the legal rights of the copyright holders of prior works? In this regard, the judge of the Beijing No. 4 Intermediate People's Court believed that it depends on whether the degree of reference and use of the AI ​​work's work on other people's works constitutes substantial similarity. If it is just a study and imitation of other people's creative techniques, rules, techniques, and styles, it is difficult to identify it as "plagiarism." However, if it is a mechanical copying and splicing of other people's works, it may constitute infringement. Therefore, on the basis of using AI to obtain inspiration and creativity in the early stage, creators should inject their own ideas, perspectives, humanistic emotions, etc. to substantively change the content and make the ownership and subsequent use of the work more secure.

2. AI “face-changing” may infringe on portrait rights

The iteration of technology has made many impossible things possible, as is the case with the "AI Stefanie Sun" that has exploded on the Internet recently. Many creators use Stefanie Sun's timbre to train AI and use synthesized timbres to cover the works of other singers. Stefanie Sun, who has not released an album for many years, instantly has more than a thousand cover songs. Even she herself said that "human beings cannot surpass it." Just around the corner." So, are these AI covers legal?

In fact, the singer's singing voice and timbre itself do not fall within the scope of protection of copyright law. However, Article 1023 of the Civil Code stipulates that the protection of natural persons' voices shall be governed by the relevant provisions on the protection of portrait rights, that is to say , shall not use information technology means to forge to infringe other people's rights to sound.

In addition, some people claim exemption on the grounds of "for entertainment only and not for commercial use". However, according to the provisions of Article 42 of the Copyright Law, sound recording producers use musical works that have been legally recorded as sound recordings by others to make sound recordings. You may do so without the permission of the copyright owner, but you must pay remuneration in accordance with regulations. The disclaimer is not a reason to prevent copyright infringement. Moreover, some people will receive platform revenue or user rewards after uploading AI cover works. Although they do not have to pay to listen, the producers do gain income from this, so it is difficult to define their works as fair use under the copyright law. "Free show".

It can be seen that although AI technology is good, it must be guarded against abuse. There was a time when celebrities such as "Yang Mi" and "Reba" frequently appeared in the live broadcast room to bring goods, and sharp-eyed netizens saw through it at a glance. It turned out that these were AI face-changing technologies, and some people tried to wear a "star face" Attract more users to enter the live broadcast room. In doing so, although there is traffic, it also infringes on the legitimate rights and interests of others. According to Article 1019 of the Civil Code, no organization or individual may infringe on the portrait rights of others by vilifying, defacing, or using information technology means to forge. Without the consent of the portrait right holder, the portrait of the portrait right holder shall not be produced, used, or disclosed, unless otherwise provided by law. Without the consent of the portrait right holder, the right holder of the portrait work shall not use or disclose the portrait of the portrait right holder in any manner such as publishing, copying, distributing, renting, or exhibiting. The portraits of celebrities have more commercial value than ordinary people. Illegally "transplanting" celebrity portraits onto anchors is equivalent to obtaining celebrity effects at zero cost. It not only infringes on the portrait rights of the celebrities whose faces have been changed and the economic interests they represent, but also affects Other operators who pay to use celebrity portraits in accordance with the law are also obviously unfair.

In addition, some netizens have uploaded their own photos and replaced the faces of actors in film and television dramas with their own faces. This behavior also carries legal risks. According to the legal provisions of portrait rights, "reverse face swapping" is also an act of infringing on other people's portrait rights by forging information technology means. In particular, face-changing characters in film and television dramas and uploading videos to the Internet may also constitute an infringement of copyrights such as the right to disseminate film and television information online and the right to protect the integrity of the work. The relevant copyright holders may reserve the right to pursue their liability and obtain compensation.

The judge of the Beijing No. 4 Intermediate People’s Court reminded that “face-changing” may also trigger other disputes over personality rights. A video blogger recently used AI face-changing technology to replace the face of the male protagonist in an intimate scene in a film and television drama with his own, causing discomfort to many netizens and fans of the actor. The actors' intimate scenes in the play are performance needs, which are reflected in the contract and script. They are a necessary interpretation of the plot based on the actor's knowledge and consent. However, the blogger’s behavior not only constituted rudeness and offense to the actress, but also may be considered as “violating another person’s will by means of language, writing, images, physical behavior, etc.” as stipulated in Article 1010 of the Civil Code. "Others commit sexual harassment" and bear civil liability. What's more, in order to "create pornographic rumors" or illegally make profits, the portraits of female celebrities or ordinary women are changed into pornographic videos, which seriously infringes on the reputation rights of others and causes great mental pain to the parties concerned. They should be punished according to law. punish.

3. Focus on preventing the use of AI to commit crimes

While people are still exploring the various conveniences of AI, criminals have already focused on this technology. According to reports, Mr. Guo’s WeChat “friend” from Fujian contacted him via video and said he wanted to borrow a deposit of 4.3 million yuan. Based on his trust in the video call, Mr. Guo completed the transfer within 10 minutes and soon discovered that he had been cheated. It turns out that the scammer first stole the WeChat account of Mr. Guo's friend, learned about Mr. Guo's financial strength, and then obtained the "friend's" face, voice and other information for later synthesis, and finally completed the fraud, making it difficult to detect. Originally, remote communication was created for convenience, but the misuse of AI technology has reduced people's trust in online communication. Just imagine, if someone uses AI technology to produce illegal videos, fake news, etc., this will undoubtedly be detrimental to the social trust system.

Integrity is crucial to ensuring transaction security, but improper use of deep synthesis technology may increase the risk of default, especially in the field of transactions based on the intellectual achievements of the other party. When the creator's work is text or pictures generated through AI tools such as ChatGPT and Midjourney, the originality of the work may be flawed. For example, in terms of copyright licensing or transfer contracts, unless both parties to the contract have reached an agreement on the use of AI for creation, such synthesized works may constitute defective performance due to insufficient originality, or may have rights defects due to infringement of other people's copyrights, giving the licensee The party or the transferee may cause unnecessary losses in the subsequent commercial use of the work.

In addition, the accessibility of AI technology will also increase moral hazard. The judge of the Fourth Intermediate People's Court of Beijing said that compared with the workload required for individuals to carry out independent intellectual work, the time, economic and energy costs are far different between the workload required to complete work using AI. If writing, painting, etc. Practitioners in the field rely heavily on AI to complete work without informing paying customers, and still quote according to the standards completed independently by individuals. The order of healthy competition in related industries will be broken, and professional ethics will also be in crisis. Of course, there are also academic fields where the stakes are high. It is nothing new for students to use artificial intelligence tools such as ChatGPT to assist in writing papers. However, at this stage, papers created by AI still have many shortcomings, such as lack of novelty, logical expression, and even references. It may be that AI has fabricated it randomly. Not only does such academic achievement have no value at all, it actually aggravates the phenomenon of academic misconduct.

In the face of related risks, we must not only make remedial measures, but also take precautions. The "Internet Information Services Deep Synthesis Management Regulations" promulgated by the Cyberspace Administration of China were officially implemented on January 10 this year. They mainly stipulate the obligations of deep synthesis service providers, which require deep synthesis service providers to provide intelligent dialogue and synthesized human voices. , face generation, immersive simulation scenes and other services that generate or significantly change the function of information content should be clearly marked to avoid confusion or misunderstanding by the public. It can be seen that to prevent risks, we must first start with technology developers and service providers.

For individuals, the protection of biometric information such as faces, fingerprints, and voices, as well as personal information such as ID cards, bank cards, and social media accounts, has become increasingly important. If an acquaintance requests a transfer online, their identity must be confirmed through multiple verifications. This is especially true for the elderly who do not know much about AI technology.

Photo courtesy: Visual China

Source: Beijing Daily Client Reporter: Yuan Jing

Process Editor: u060

The above is the detailed content of Judge reminds - Be wary of the legal risks of misuse of AI technology. For more information, please follow other related articles on the PHP Chinese website!