生成式人工智能的法律风险及侵权责任认定  

Legal Risks and Infringement Liability Determination of Generative Artificial Intelligence

作　　者：严雅茹 骆小春[1] YAN Yaru;LUO Xiaochun(School of Law and Politics,Nanjing Tech University,Jiangsu,Nanjing,211816,China)

机构地区：[1]南京工业大学法政学院,江苏南京211816

出　　处：《盐城师范学院学报（人文社会科学版）》2024年第2期54-65,共12页Journal of Yancheng Teachers University（Humanities & Social Sciences Edition)

摘　　要：生成式人工智能迅猛发展的背后折射出科技进步与法律制度发展的辩证关系。当前生成式人工智能技术带来众多科技红利的同时,也带来诸多法律风险,主要存在于个人信息保护、生成内容的版权问题等方面,并由此带来侵权责任制度的选择问题。秉承鼓励创作与创作者权益保护的理念,应当认可一定情形下生成式人工智能生成内容具有可版权性,并为版权权益归属、版权侵权情形下侵权主体的界定提供依据。综合考量各主体的风险控制能力与安全防范义务,生成式人工智能的侵权责任应当采取一般侵权责任制度并辅以过错推定规则。基于生成式人工智能侵权的特殊性,应当对服务提供者变通适用“通知删除”规则。The rapid development of generative artificial intelligence reflects the dialectical relationship between technological progress and the development of legal system.The current generative artificial intelligence technology brings numerous technological dividends,along with many legal risks,mainly in the fields of personal information protection,copyright issues of generated content,and the selection of infringement liability determining systems.Adhering to the principle of encouraging creativity and protecting the rights and interests of creators,we suggest that the content created by generative artificial intelligence should be recognized as copyrightable under certain circumstances,which provides a basis for the ownership of copyright and the determination of infringing parties.Taking into account the risk control capabilities and security obligation of each party,we suggest that we should adopt a general infringement liability determination system,supplemented by fault presumption rules,in consideration of the infringement liability of generative artificial intelligence.Based on the particularity of infringement of generative artificial intelligence,we should flexibly apply the“notice-and-takedown regime”on service providers.

关 键 词：生成式人工智能 可版权性 产品责任 过错责任原则 过错推定 

分 类 号：D923.1[政治法律—民商法学]