人工智能要素的政府供给:法理基础与制度因应  

On Governmental Supply of Artificial Intelligence Elements:Jurisprudential Basis and Institutional Response

作　　者：苏宇[1] Su Yu

机构地区：[1]中国人民公安大学法学院

出　　处：《财经法学》2025年第2期77-93,共17页Law and Economy

基　　金：2024年度教育部人文社会科学研究青年基金项目“生成式人工智能的公私协同治理机制研究”(24YJC820035)的阶段性成果。

摘　　要：数据、算力与算法三要素对人工智能发展具有重要支撑作用。人工智能三要素均不同程度上具有公共物品属性,政府供给人工智能要素具备法理上的正当性基础。公共物品属性存在基于分享潜力效应和区分受益机制的两种消解途径,但非公共物品也在市场失能、市场缺位和资源依赖三种例外情形下需要政府供给。数据、算力与算法的公共物品属性和政府供给需求各不相同,政府供给机制亦须适应此种差别。法律应对政府供给机制作差异化授权,厘定政府供给人工智能要素的政策空间,并就政府供给的退出机制预作规定。The triad of data,computing power,and algorithms plays a critical supporting role in the development of artificial intelligence.To varying degrees,these three elements of AI possess attributes of public goods,providing a legitimate foundation for governmental supply of AI elements.The public goods characteristics can be mitigated through two pathways based on the potential for sharing effects and differentiated beneficiary mechanisms;however,non-public goods may also require governmental supply under exceptional circumstances such as market failure,market absence,and resource dependency.Given the distinct public goods attributes and varying demands for governmental supply associated with data,computing power,and algorithms,it follows that governmental supply mechanisms must be adaptive to these differences.Legislation should authorize differential mechanisms for governmental supply,define the policy space for governmental supply of AI elements,and establish exit mechanisms for governmental supply.

关 键 词：人工智能 公共物品 政府供给 政策空间 

分 类 号：D927[政治法律—法学] D922.17[自动化与计算机技术—控制理论与控制工程] TP18[自动化与计算机技术—控制科学与工程]