生成式人工智能数据风险的法律保护与规制研究——以ChatGPT潜在数据风险为例

Research on Legal Protection and Regulation of Generative AI Data Risks:the Case Study of ChatGPT’s Potential Data Risks

ChatGPT是美国人工智能研究实验室OpenAI推出的一种人工智能技术驱动的自然语言处理工具,其深度的学习能力和自我纠正能力,提高了人工智能与人类之间的交往。ChatGPT在实现科学技术革新的同时,也存在敏感数据泄露的风险。ChatGPT在保护个人主体数据、企业主体数据和国家主体数据方面需要法律予以规范。因此,如何应对ChatGPT带来的数据安全,成为生成式人工智能亟待解决的问题。为此,通过分析ChatGPT运作阶段和数据主体应用现状可能存在的数据风险,以我国数据规制法律框架为基础,借鉴域外人工智能立法模式,探究ChatGPT的数据风险法治化应对策略,以此合理构建我国生成式人工智能的数据风险法律规制体系。

ChatGPT is an AI technology-driven natural language processing tool introduced by OpenAI,an AI research laboratory in the U.S.Its in-depth learning and self-correcting capabilities have improved the interaction between AI and humans.ChatGPT,while realizing scientific and technological innovations,has the risk of leakage of sensitive data.ChatGPT’s ability to protect personal subject data,corporate subject data and state subject data needs to be regulated by law.Therefore,how to deal with the data security issues brought by ChatGPT has become an urgent problem for generative AI intelligence.To this end,this paper analyzes the data risks that may exist in the operation stage of ChatGPT and the application status quo of data subjects,and based on the legal framework of data regulation in China,draws on the foreign legislative model of AI to explore the response with rule of law to data risks of ChatGPT,so as to reasonably construct the data risk legal regulation system of generative AI in China.