试论人工智能生成内容的合理使用

On the Fair Use of AI-Generated Content

摘要: 近年来，人工智能技术得到了飞速的发展，已经能够独立生成内容，与人类作品之间的差距微乎其微，已经初步符合了作品客体的要求，需要对其进行赋权保护。同时，为避免权利滥用需对其作出一定限制，即关于合理使用制度能否适用于人工智能生成内容。本文第一部分首先进行了相关立法现状以及人工智能生成内容的发展铺垫，第二、三部分分别从理论支撑、必要性展开论述。第四部分提出了关于人工智能合理使用的认定标准，结合“额头出汗原则”进行宽容化的认定和原则性的规定，同时制定客观具体的规章文件以使认定标准具有可操作性。第五部分得出结论，对于人工智能生成内容，既然需要赋权保护那么就应当适用合理使用制度，以避免权利失衡，同时相关的文件也表明了这一倾向。

Abstract: In recent years, artificial intelligence technology has developed rapidly, and it has been able to generate content independently, and the gap between it and human works is minimal, which has initially met the requirements of the object of the work and needs to be empowered and protected. At the same time, in order to avoid abuse of rights, it is necessary to limit whether the fair use system can be applied to AI-generated content. The first part of this paper first paves the way for the current situation of relevant legislation and the development of AI-generated content, and the second and third parts discuss the theoretical support and necessity respectively. The fourth part puts forward the identification standards for the reasonable use of artificial intelligence, combines the “forehead sweating principle” for tolerant identification and principled provisions, and formulates objective and specific regulatory documents to make the identification standards operable. Part V concludes that for AI-generated content, since empowerment protection is required, fair use regimes should be applied to avoid rights imbalances, and the relevant documentation indicates this tendency.