人工智能生成物的出现导致知识产权法理论基础遭遇困境，既有规则存在解释与适用困境。生成式人工智能技术带来的知识产权法律争议背后是过程论抑或结果论的不同选择。化解争论与分歧的根源在于对生成式人工智能涉及的对象予以准确、科学的界定，对相关客体的关照应从知识产权转变为更广泛的数据要素视角。数据要素视角下，传统的劳动价值说、个人自由主义与激励理论应进行时代性解读，以适应人工智能科技发展与利益平衡的现实需求。结合数字经济与产业发展动态、知识产权法的社会功能以及保护范围扩张趋势，对生成式人工智能技术创新与应用需作人本主义的解读，推动AIGC（ArtificialIntelligenceGenerativeContents，人工智能生成内容）知识生产与理性运用的利益协调，基于数字市场中知识产权与数据要素的复杂关系，开展制度价值判断与效果评估。

The emergence of AI-Generated Contents (AIGC) causes theoretical dilemma of intellectual property law. Existing rules have difficulties in interpretation and application. The definition of intellectual property subject in existing laws is mainly based on established facts,and the emergence of AIGC changes the creation process of intellectual property, shakes the long-standing philosophical foundation of intellectual property system, and also brings difficulties to the judgment in judicial practice.
In the application of AIGC, the devices need to learn, mine, and process a large number of existing works and materials, then automatically calculate and generate contents through algorithmic models according to natural language instructions. In this process, the primary issue is whether AIGC should be classified as intellectual property subject matter. According to the traditional rules and regulations, the subject matter of intellectual property should be human intellectual output. However, human intellectual involvement is really limited in AIGC, while the appearance of AIGC looks like human intellectual output. The opinions on whether the AIGC should be protected under intellectual property law are divided into two categories. Behind this debate are the different choices of process perspective and result perspective.Nowadays, the scope of intellectual property protection shows an obvious trend of expansion, through extensive interpretation of traditional subject matter, or the introduce of new subject matter to the law, to realize inclusive and extended arrangements. As a result, there are also two kinds of legal attitudes towards the development of artificial intelligence, one is to expand the scope of traditional subject matter to cover new things, the other is to design new rules to solve new problems.
In this case, the start point of resolving disputes and disagreements is to accurately and scientifically define the subject matter involving generative AI, and the analysis position should be shifted from intellectual property to a broader perspective of data elements. The current data intellectual property registration’ system in China may be an alternative. From the perspective of data elements, the traditional labor theory of value, individual liberalism and incentive theory should be interpreted to meet the realistic needs of the development of artificial intelligence technology and the balance of interests. Both the technical nature and the technical process show that AIGC is actually simulating human activities of information and data extraction, learning and thinking. The output of AI is essentially the result of information and data processing. Old rules and ideas cannot support new practices, and social costs and historical trends should be taken into account in making new institutions and rules. Combined with the trends of digital economy and industrial development, the social function of intellectual property law and the expansion of intellectual property protection, it is necessary to make a humanistic interpretation, promote interests coordination, facilitate AI knowledge production and rational application, and carry out institutional value judgment and effect evaluation based on the complex relationship between intellectual property and data elements in the digital market.
Considering that the data rules are still in the turnaround zone and the inherent rules of intellectual property need to be adjusted in time, it is essential to comprehensively consider the connection and interaction between the data rules and the intellectual property system to find a solution to the legal issues of AIGC. From the perspective of future data governance, AIGC rules should be constructed according to the separation-of-rights framework rather than limited to the traditional intellectual property system. The definition of rights in AIGC cannot be separated from the characteristics and ownership issues of the data elements.This will help promote the comprehensive governance of artificial intelligence in the future and the long-term development of the digital economy.
In order to balance the benefits brought about by technological innovation and application effectively, the reasonable choice is to break through the boundaries of traditional information carriers such as works, patents and trademarks, and pay attention to the distribution of benefits involved in various fields such as data collection, protection, processing, transaction, circulation and use. It is necessary to explore the balance between private and public interests and establish a coordination mechanism based on the ‘data property rights’ framework so as to provide a legal basis for the governance of artificial intelligence and the development of the domestic digital economy.