The Beijing Internet Court recently held a public hearing in accordance with the law and heard an "AI Vincent Picture" copyright case. The case was not pronounced in court.
In this case, the plaintiff, Mr. Li, is an online self-media creator. He generated the pictures involved in the case through an online open source software and published them on the Internet with the title "The Tenderness Brought by the Spring Breeze". Later, Mr. Li discovered that in an article published by the defendant Ms. Liu, the pictures generated by him were used without authorization, and Mr. Li's account signature watermark was also cut off. Mr. Li believed that his information network dissemination rights and signature rights had been infringed, and he went to court to require the defendant Ms. Liu to compensate him for economic losses of 5,000 yuan and apologize.
The trial process of the case was broadcast live by the China Central Radio and Television Station News Center. Professor Cui Guobin, director of the Intellectual Property Law Research Center of Tsinghua University Law School, and Huang Yingyuan, a computer image algorithm engineer, were guests of the program and discussed the technical and legal issues involved in AI-generated content. discussed. The program was broadcast simultaneously on the CCTV news client, new media platforms and the official Douyin, Kuaishou and Video accounts of the Beijing Internet Court, attracting the online attention of 170,000 netizens.
Neither party has any objection to the fact that the defendant used the pictures involved in the case to illustrate the case. The defendant Ms. Liu stated that she did use the picture as an illustration in the article, but she had no intention of infringement and no other commercial activities. After receiving the notice, Ms. Liu deleted the relevant pictures.
The reporter learned from the court hearing that there are three main points of dispute in this case: first, whether the pictures claimed by the plaintiff to infringe constitute a work and what kind of work they belong to; second, whether the plaintiff enjoys the copyright of the pictures involved; third, whether the alleged infringement constitutes infringement and what kind of infringement it is. Assumption of responsibility. According to the relevant provisions of my country's Copyright Law, in order to determine whether a work can be recognized as a legally protected work, it is first necessary to determine whether it has "original" expression. During the trial, both the plaintiff and the defendant also presented evidence in this regard.
Before discussing the focus of the dispute in this case, a more basic issue must be clarified - should the pictures generated by AI be regarded as the work of AI or the work of humans? This is extremely important for determining whether the defendant in the case has infringed the copyright. This is because AI is not an "author" recognized by my country's Copyright Law. Similarly, in March this year, the United States Copyright Office (USCO) issued a new guideline on the copyright of AIGC works, explaining the copyright conditions of AI works. It is mentioned in the guidelines that works directly generated by AI tools such as ChatGPT are not protected by copyright law, and only works created by humans using AI as an auxiliary tool can be supported by copyright. The guidance specifically states that copyright can only protect works created by humans and that the term "author" used in the Constitution and copyright law excludes non-human beings.
Back to the question itself, is the author of the AI drawings a human or an AI? Instead, it needs to be analyzed on a case-by-case basis. This is because the concept of “AI generated images” itself is extremely complex. Take the Stable Diffusion software involved in this case as an example. The user can not participate in the AI drawing process at all. Just click "One-click Drawing" with the mouse, and the software will spontaneously generate a beautiful picture. At this time, the software user Without condensing certain intellectual achievements in the picture, it is difficult to say that the author is the author of the picture. But in more cases, the user actually plays the role of the "drawer" of the picture. By inputting keywords into the software, the user can guide all the details of the AI painting. At this time, the selection of keywords, the control of the style of the generated pictures, etc. all condense the intellectual achievements and labor achievements of the users behind the AI image generation software, so the judgment of the "author" at this time may be different from the first case. Are not the same.
Regardless of the final outcome, the verdict of this case will have a huge impact on my country's AI imaging industry. The issuance of various normative documents for AI generation technology in the future is the general trend of the times, and more and more cases like today will appear in the future. A case involving copyright disputes over AI-generated works.