摘要
生成式人工智能的技术跃进架空了个人信息处理的告知同意规制和最小必要原则,引发了虚假信息生成和个人信息泄漏的广泛风险迭代问题。传统个人信息的权利保护路径面临认知和结构困境,无法应对生成式人工智能给个人信息保护带来极大挑战。以风险控制为导向的个人信息保护机制不强调信息主体对个人信息的绝对控制,旨在通过识别、评估、分配和管理将风险控制在最小范围内,可以灵活和实用地平衡生成式人工智能应用中的信息利用和风险控制,提供有效的解决方案。在风险控制理念下,对告知同意规则和最小必要原则进行风险化解释与调试,并建立从预防到识别再到控制的虚假信息生成风险的全过程应对机制,以及基于风险的个人信息保护合规管理体系,是当前的最优选择。
The technological leap forward of generative artificial intelligence overcomes the regulation of consent and the principle of minimum necessity in personal information processing,and leads to the extensive risk iteration problems of false information generation and personal information leakage.The traditional way of protecting the rights of personal information is faced with cognitive and structural difficulties and cannot cope with the great challenges brought by generative artificial intelligence to the protection of personal information.The risk controloriented personal information protection mechanism does not emphasize the absolute control of the information subject over the personal information,but aims to control the risk in a minimum range through identification,assessment,distribution and management.It can balance the use of information and risk control flexibly and practically,and provide effective solutions.Under the concept of risk control,it is the best choice at present to interpret and debug the rules of consent and the principle of minimum necessity,and establish a whole-process response mechanism from prevention to identification to control of the risk of false information generation,as well as a risk-based personal information protection compliance management system.
作者
钭晓东
Dou Xiaodong(Guanghua Law School of Zhejiang University,Hangzhou Zhejiang 310088)
出处
《政法论丛》
北大核心
2023年第4期59-68,共10页
Journal of Political Science and Law
基金
台达集团中达环境法学教育促进计划
“党的二十大精神研究阐释”专项课题《数字赋能生态治理中国式现代化及其法治保障》获中央高校基本科研业务费专项资金的资助
国家社科基金重大招标项目“加快推进生态治理体系与治理能力现代化研究”(编号20&ZD091)的阶段性成果。
