行为解读与利益平衡:人工智能训练的著作权合理使用

Interpretation of Behavior and Balance of Interests:Fair Use of Copyright in Artificial Intelligence Training

人工智能在开发阶段面临训练数据来源的著作权合法性问题。在制度层面上,体现为作品使用新方式的出现对传统合理使用规则的冲击问题;在法理层面上,体现为著作权人与公众的利益分配问题。鉴于人工智能的数据使用行为是非传统复制行为,考虑到人工智能产业发展对于大规模训练数据的实际需求与确保公众对作品的适当接近、促进文化艺术繁荣的公共利益需求,有充分理由将人工智能使用作品进行训练的行为视为一种转换性使用行为,从而纳入合理使用范畴。同时,为作者提供配套的名誉补偿机制及人工智能生成内容实质性侵权的救济路径;倡导企业建立共享数据资源库以达成企业经济效益与社会责任的平衡。

Artificial intelligence faces the legality of training data sources during its development.At the institutional level,this manifests as the challenge of new forms of work usage to traditional fair use rules;from a jurisprudential perspective,it manifests as the issue of interest allocation between copyright holders and the public.Considering that the data usage behavior of artificial intelligence is non-traditional replication behavior,and taking into account the actual demand for large-scale training data in the development of the artificial intelligence industry,as well as the need to ensure the public’s appropriate access to works and promote cultural and artistic prosperity,there are sufficient reasons to consider the behavior of using works for artificial intelligence training as a transformative use,thereby including it in the scope of fair use.At the same time,it is necessary to provide authors with a corresponding mechanism for reputational compensation and a remedy for substantial infringement of artificial intelligence-generated content;advocate for enterprises to establish shared data repositories to achieve a balance between corporate economic benefits and social responsibilities.