生成式人工智能著作权侵权的规制路径研究

Research on the Regulatory Paths of Copyright Infringement by Generative Artificial Intelligence

摘要: 基于生成式人工智能的运行逻辑，其在输入阶段的数据训练以及输出阶段的内容生成，均会在著作权领域产生一定的侵权风险。而目前的法律规范，大多过于笼统，使得法官在实践中难以依据明确的裁判规则高效审理，形成难以同案同判的司法现实，亟需相应的规制路径加以优化。以问题导向为出发点，通过对生成式人工智能著作权侵权进行二元类型化分析，进一步明确生成物的著作权归属及责任分配规则，并从规制算法黑箱、统筹专项立法、构建法定许可制度、强化源头防控等角度进行路径构建以适应当前司法实践的客观需要。

Abstract: Based on the operational logic of generative artificial intelligence, both its data training in the input phase and content generation in the output phase pose certain infringement risks in the field of copyright law. However, most current legal norms are too general and vague, making it difficult for judges to efficiently adjudicate such cases based on clear adjudicative rules in judicial practice. This has resulted in the judicial reality of inconsistent adjudication of similar cases, and corresponding regulatory approaches are urgently needed for optimization. Taking a problem-oriented approach as the starting point, this paper conducts a binary typological analysis of copyright infringement by generative artificial intelligence, further clarifies the copyright ownership of generated content and liability allocation rules, and constructs regulatory paths from the perspectives of regulating algorithmic black boxes, coordinating special legislation, establishing a statutory licensing system, and strengthening source prevention and control, so as to meet the objective needs of current judicial practice.