人工智能辅助生成影视内容的版权困局与纾解之道——以Sora为视角

Copyright Dispute and Institutional Relief of AI Aided Generated Film and Television Content--Take Sora as an Example

文生视频大模型Sora的诞生给影视产业带来了巨大的变革和挑战,借助人工智能模型生成的影视内容的可版权性认定正深陷困局。理论层面,从著作权主体视角来看,人工智能模型不具备成为作者的法理依据,尚无法取代人类在影视创作中的主体地位;从著作权客体视角来看,人工智能辅助生成内容已然满足“以作品为中心”的客观标准。实践层面,通过引入标识化管理的制度设计,将有助于人工智能辅助生成影视内容管理工作提质增效。今后应明确人工智能辅助生成内容上的标识不构成“署名”,进一步更新升级数字水印技术,持续完善标识化管理规则体系建设,多管齐下确保发挥文生视频技术的积极效应,推动人工智能朝着科技向善的方向发展。

The birth of text to video foundation model Sora has brought great changes and challenges to the film and television industry,and the copyrightability identification of artificial intelligence(AI)aided generated film and television content is in deep trouble.At the theoretical level,from the perspective of the subject of copyright,AI model does not have the legal basis to become an author,and cannot replace the subject status of human beings in film and television creation;From the perspective of the object of copyright,AI aided generated content already meets the objective standard of"Work Centered".At the practical level,through the introduction of the system design of identification management,it will help improve the quality and efficiency of AI aided generated film and television content management.In the future,it should be clear that the logo on AI aided generated content does not constitute a""Signature",further update and upgrade the digital watermarking technology,continue to improvethe construction of the identification management rule system,and ensure that the positive effect of text to video technology is played in a multipronged approach to promote the development of AI towards the direction of technology for social good.