风险、困境与对策:生成式人工智能带来的个人信息安全挑战与法律规制

Risks,Dilemmas and Countermeasures:Personal Information Security Challenges and Legal Regulation Caused by Generative Artificial Intelligence

以ChatGPT为代表的生成式人工智能产生智能涌现,具有自主学习性、强交互性、创新性等特点。为规制其个人信息安全风险,从生成式人工智能的数据收集与处理、算法运行与其生成性内容三个要素为切入点,需要对个人数据侵权、算法偏见、生成性内容的非法利用引发个人信息安全风险进行分类讨论。结合国内外已有的人工智能立法实践,生成式人工智能个人信息规制存在法律规范缺位、规则适用、治理手段乏力等现实困境,仍需要从立法、监督、司法救济三个层面提出相应的法律规制建议,以促进生成式人工智能个人信息法律治理体系的完善。

The generative AI represented by ChatGPT generates intelligent emergence with features such as autonomous learning,strong interactivity,and innovativeness.In order to regulate its personal information security risk,from the data collection and processing,algorithm operation and its generative content of generative AI three elements as an entry point,it is necessary to classify and discuss personal data infringement,algorithmic bias,and illegal use of generative content triggering personal information security risk.Combined with the existing AI legislative practices at home and abroad,the regulation of personal information of generative AI has the real dilemma of the absence of legal norms,the application of rules,the lack of governance means,to mention just a few.It is still necessary to put forward the corresponding legal regulation proposals from the three levels of legislation,supervision and judicial relief,in order to promote the improvement of the legal governance system of personal information of generative AI.