Artificial intelligence brings copyright disputes that need guidance from legal policies
2025-01-17
The global artificial intelligence technology is rapidly changing, and generative artificial intelligence technology represented by ChatGPT is changing the world, with copyright being the most affected. Artificial intelligence cannot be created out of nothing. Massive Internet information is the fertile ground for its cultivation. This means that it will inevitably fall into controversy over intellectual property and content plagiarism. In recent years, conflicts between content copyright and artificial intelligence have occurred frequently. In the just passed year of 2024, there have been influential and highly anticipated cases both domestically and internationally. In the "2024 Digital Copyright Protection and Development Annual Key Words" recently released by the National Copyright Trade Base of Renmin University of China, "AI big model corpus training copyright challenge" ranks the third among the eight key words. Finding a balance between protecting the rights and interests of creators, stimulating creative enthusiasm, and promoting industrial innovation has become a major issue that must be faced at present. Copyright protection is facing challenges. In recent years, the development of large-scale models has been surging and surging. Looking at the world, apart from ChatGPT, AI models such as Midjourney, Stable Diffusion, and Sora have become popular, and the artificial intelligence industry is rapidly developing in technologically advanced countries. China is no inferior. Many Internet companies have entered the market and launched artificial intelligence models such as ERNIE Bot, Tongyi Qianwen and Doubao. The emergence of these products brings obvious convenience, making copywriting, information retrieval, data processing, and more efficient, profoundly affecting the traditional industry order and work habits. Xiaofei is a civil servant in a company. She told a reporter from the Legal Daily that she often turns to the above-mentioned products for help when she has any difficult questions. "Especially when preparing formatted text materials, she explains the requirements as detailed as possible and usually presents content that meets the requirements, saving me a lot of time in collecting materials and understanding relevant background knowledge in the past. Professor Lu Xianghua from the Department of Information Management and Business Intelligence at Fudan University School of Management is also an active user of big models. In Lu Xianghua's view, the application of large models can save at least 70% of energy compared to the past. The popularity of artificial intelligence models has also sparked controversy. For example, in 2024, a certain online novel platform notified contracted authors to sign a supplementary agreement allowing their contracted works to be used for AI model training, which sparked opposition and later deleted the clause. But the doubts have not disappeared. Some authors have expressed concerns that they may lose control over their works or that unpublished materials may be absorbed by AI and output similar content first, making them 'plagiarism infringers'. The music industry also has similar concerns. Looking back on the past year, many creators have become proficient in using AI tools to create music content. As long as they input simple text prompts such as music style and genre, they can get a complete song in just a few seconds. At the 2024 Digital Copyright Protection and Development Forum recently held by the National Copyright Trade Base of Renmin University of China, Duan Yuping, the former deputy director of the Copyright Administration Bureau of the Propaganda Department of the CPC Central Committee, said frankly that new technologies have indeed brought new challenges to music copyright protection, and specifically mentioned the example of AI Sun Yanzi causing copyright disputes. Driven by new technologies such as artificial intelligence, the scope of works is expanding day by day, and the boundaries of copyright are gradually blurring. Traditional copyright protection methods are becoming increasingly difficult to cope with the challenges of new technologies, new formats, and new models Said Zhi Xiaoqiang, member of the Standing Committee of the Party Committee and Vice President of Renmin University of China. Litigation cases involve conflicts between copyright and artificial intelligence in multiple fields, and are not always presented in the form of "debates". There are also multiple cases that have been brought to court. Li Fangli, deputy director of the National Copyright Trade Base of Renmin University of China, told reporters that there are relevant judicial practices at home and abroad. As for foreign countries, there have been multiple lawsuits against AI big model training data since The New York Times sued OpenAI and Microsoft for copyright infringement. The reporter found that on December 27, 2023, The New York Times announced that it would file a lawsuit with the Southern District Federal Court of New York, accusing Microsoft and OpenAI of copyright infringement and abuse of the media's intellectual property. Subsequently, three digital news media outlets in the United States, The Intercept, Raw Story, and AlterNet, filed copyright infringement lawsuits against OpenAI on February 28, 2024. In June 2024, artificial intelligence startup Perplexity was accused of content plagiarism by Forbes and tech media Wired, claiming to have found their reports in Perplexity's Pages tool. On October 21st, Dow Jones, a subsidiary of media mogul Rupert Murdoch's News Corporation, and The New York Post filed a lawsuit against Perplexity, alleging that they "illegally copied a large amount" of their copyrighted works. The music industry is also not stopping. In June 2024, the Recording Industry Association of America sued two AI music companies, Suno and Udio, accusing them of infringing music copyrights while training AI tools. In addition, there are similar cases in Canada, Germany, and other countries. Jin Haijun, a professor at the Law School of Renmin University of China, said that he had counted about a dozen similar lawsuits, which troubled copyright works in various fields such as text, pictures and music. There are also relevant cases in China. On February 8, 2024, the Guangzhou Internet Court made the world's first judgment on the tort liability of the generative AI platform, finding that the defendant AI company violated the plaintiff's right to copy and adapt Altman's works. On April 23 of the same year, the first case of AI generating voice personality right infringement in China was decided by the Beijing Internet Court in favor of the voiceover. AI developers used the voice of the voiceover and developed AI text to voice products involved in the case without legal authorization, which constituted infringement. On June 20 of the same year, the Beijing Internet Court held an online court session to hear four cases of painters suing AI painting software developers and operators for copyright infringement, which was the first case involving AI painting model training copyright infringement in China. Commercial use is particularly complex. Currently, there are no clear provisions in China's copyright law regarding the protection of AI generated products; There is still controversy over whether the copyright of AI generated content belongs to the developers or actual users of AI technology, and clear legal and policy guidance is needed. The public is actively taking action. On December 4, 2024, the Chinese Character Copyright Association proposed that the construction of generative artificial intelligence corpora should ensure the legality of data sources. The development of artificial intelligence technology should not only protect the legitimate rights and interests of creators and content producers, but also stimulate the innovation and creativity of the whole society, and promote the high-quality development of the industry. On December 16, 2024 local time, the "Artificial Intelligence Creative Rights Alliance" was officially established in the UK, proposing three key principles on copyright and generative AI policies, and calling on the UK government to use these three principles as a framework for formulating AI policies. The above-mentioned large models require sufficient corpus to train AI models, and during the training process, a large number of copyrighted works must be used. This involves how we use the work, or in other words, how to use the work reasonably Li Fangli said. Jin Haijun believes that in terms of improving relevant systems, special attention should be paid to two parts of the AI generated content generation process. One is the stage of output, that is, whether the content generated by AI can be protected by copyright, namely the so-called work issue and author qualification issue; Secondly, in the input stage, when using the corpus, if it involves copyrighted works, does it constitute infringement. If it constitutes infringement, can reasonable use be used as a defense. From the current international situation, the determination of whether a work constitutes a work within the meaning of copyright law is generally quite strict. When using works on the input end, the requirements are relatively relaxed. But there is a condition that it is non-commercial when used in text mining and data training. Once it enters the commercial system, it cannot be simply said that it constitutes infringement or fair use Jin Haijun said. Yan Xiaohong, Chairman of the China Copyright Association, believes that from a technical perspective, when using copyrighted works, relevant information should be disclosed if the technical conditions permit, "so that others know what you have used". In principle, it should be considered that these uses are feasible, and obtaining permission for each work is actually impossible. Guided by the "Opinions of the Central Committee of the Communist Party of China and the State Council on Building a Data Infrastructure System to Better Play the Role of Data Elements" issued in December 2022, legislation should be accelerated to clarify the principles of use. If commercial use is involved, remuneration must be paid. How to pay remuneration should be resolved through consultation between users and rights holders under the guidance of the National Copyright Administration. At the previous China Europe Symposium on Copyright Protection in the Digital Environment, the relevant person in charge of China's National Copyright Administration pointed out that China welcomes the arrival of artificial intelligence with a positive and forward-looking thinking, researches and regulates artificial intelligence, improves institutional design, explores and standardizes copyright rules related to artificial intelligence, and strengthens monitoring of the dissemination of content generated by artificial intelligence. (New Society)
Edit:Rina Responsible editor:Lily
Source:Legal Daily
Special statement: if the pictures and texts reproduced or quoted on this site infringe your legitimate rights and interests, please contact this site, and this site will correct and delete them in time. For copyright issues and website cooperation, please contact through outlook new era email:lwxsd@liaowanghn.com