从“赋权保护”到“信义保护”:生成式人工智能中个人信息保护的范式转型  

From Empowerment Protection to Fiduciary Protection:The Paradigm Shift in Personal Information Protection within Generative Artificial Intelligence

作　　者：莫旻丹 张焕培 Mo Mindan;Zhang Huanpei(School of Law and Politics,Guangdong Ocean University,Zhanjiang 524088,Guangdong,China)

机构地区：[1]广东海洋大学法政学院,广东湛江524088

出　　处：《征信》2024年第12期21-29,共9页Credit Reference

基　　金：广东省哲学社会科学规划2021年度粤东西北专项项目(GD21YDXZFX01)。

摘　　要：数字技术的飞速发展,推动着生成式人工智能的深度应用。面对生成式人工智能服务的兴起,现行个人信息保护体系所采用的“赋权保护”模式,暴露出明显的局限性,存在着静态权利预设与动态保护逻辑脱节、统一赋权逻辑与差异化保护需求相互冲突等弊端,致使用户面临权利规则落空、失范风险泛滥以及权益救济受阻等严峻挑战。对此,将制度重心转向对服务提供者行为规范的“信义保护”模式更契合个人信息保护的需要。从服务提供者与用户的身份看,二者构成一定的信义关系,将用户的信息权利作为信义保护客体具有适格性,同时亦有助于实现信息保护与数据流通利用共赢的目的。基于此,在“信义保护”模式下,应勾勒服务提供者的信义义务轮廓,形塑信义义务下的多元监督机制,拓宽背信行为下的用户救济途径。The rapid development of digital technology drives the in-depth application of generative artificial intelligence.In the face of the rise of generative AI services,the empowerment protection model adopted by the current personal information protection system has revealed obvious limitations.Problems such as the disconnection between static right presets and dynamic protection logic,the conflicts between unified empowerment logic and differentiated protection needs,have led users to face severe challenges including the failure of rights rules,an abundance of anomie risks,and obstructed rights relief.In response,shifting the institutional focus to the fiduciary protection model that regulates the conduct of service providers has become a more suitable personal information protection paradigm that meets practical needs.From the perspective of the identities of service providers and users,a certain fiduciary relationship is established between them,making users’information rights suitable as the object of fiduciary protection.At the same time,this also contributes to achieving the win-win goal of information protection and data circulation and utilization.Under the fiduciary protection model,it is necessary to outline the fiduciary duties of service providers,shape a diversified oversight mechanism under fiduciary duties,and broaden the avenues for user relief in cases of breach of fiduciary duties.

关 键 词：生成式人工智能 个人信息保护 赋权主义 信义义务 

分 类 号：F830[经济管理—金融学] D922.29[政治法律—经济法学]