人工智能风险的全流程规制逻辑及制度因应
The Full Process Regulatory Logic and Institutional Response of Artificial Intelligence Risks
胡小伟 1刘丽2
作者信息
- 1. 临沂大学法学院,山东临沂,276000
- 2. 中国政法大学法学院,北京,100091
- 折叠
摘要
人工智能风险并非确定性、单一性的风险,而是蕴含着生成多元性、演化交替性、结果未知性的体系性风险,这成为人工智能风险全流程规制的逻辑基础.全流程视域下,人工智能风险可类型化为算法设计风险、数据要素风险、平台控制风险、网络安全风险、自动决策风险等具体样态.人工智能风险规制立法应基于全流程规制视角,在规制范围上,加强全周期贯穿,实现人工智能风险的穿透规制;在规制进路上,注重全过程传导,实现人工智能风险的渐进规制;在规制内容上,强化全要素统筹,实现人工智能风险的一体规制.
Abstract
The risks of artificial intelligence are not deterministic or singular risks,but a systemic risk that contains generation diversity,evolutionary alternation,and unknown results,which constitutes the logi-cal basis for the whole process regulation of artificial intelligence risks.From the perspective of the whole pro-cess,artificial intelligence risks can be classified into specific forms such as algorithm design risks,data ele-ment risks,platform control risks,network security risks,and automatic decision-making risks.The legisla-tion on risk regulation of artificial intelligence should be based on the perspective of whole process regulation.In terms of regulatory scope,we should strengthen the full cycle penetration and achieve the penetration regu-lation of artificial intelligence risks;On the road of regulation,we should pay attention to the whole process transmission and achieve gradual regulation of artificial intelligence risks;In terms of regulatory content,we should strengthen the overall coordination of all factors and achieve integrated regulation of artificial intelli-gence risks.
关键词
人工智能风险/全流程规制/人工智能法Key words
Artificial intelligence risks/Whole process regulation/Artificial Intelligence law引用本文复制引用
出版年
2024
NETL
NSTL
万方数据