生成式人工智能对数字政府建设的变革、风险及应对  

Transformations, Risks and Responses of Generative ArtificialIntelligence in the Construction of Digital Government

作　　者：张夏恒 ZHANG Xia-heng(China Digital Trade Research Institute,Zhejiang University,Hangzhou,Zhejiang 310058)

机构地区：[1]浙江大学中国数字贸易研究院,浙江杭州310058

出　　处：《公共治理研究》2025年第2期42-53,共12页PUBLIC GOVERNANCE RESEARCH

基　　金：国家社会科学基金一般项目“跨境电商推进我国数字贸易强国建设机制与路径研究”(编号:22BJY014)。

摘　　要：数字政府建设是推进国家治理体系和治理能力现代化的重要内容,而生成式人工智能技术给数字政府建设带来诸多影响。生成式人工智能对实现数字政府行政理念从“数字化”向“数治化”更新、组织形态从“科层制”向“扁平化”重组、业务服务从“粗放式”向“精细化”升级、运行管理从“碎片化”向“合力化”优化、话语结构从“独白性”向“共识性”转变等变革具有积极作用。但也可能造成个人隐私泄露、算法黑箱歧视、权力制约失衡、过分依赖技术、商业资本入侵等问题。生成式人工智能在数字政府建设中的风险应对应从安全、技术、法律、伦理、价值五个维度入手,以发挥生成式人工智能对数字政府建设的赋能效用。The construction of digital government is an important part of promoting the modernization of national governance system and governance capacity,and generative artificial intelligence technology has brought many impacts to the construction of digital government.Generative artificial intelligence is crucial for updating the administrative concept of digital government from“digitalization”to“digital governance”,restructuring the organizational form from“hierarchical system”to“flattening”,upgrading business services from an“extensive”mode to a“refined”one,and optimizing operational management from“fragmentation”to“synergy”,transforming discourse structure from“monologue”to“consensus”.However,it may also cause problems such as personal privacy leakage,algorithmic black-box discrimination,imbalance in power restraint,excessive reliance on technology,and invasion of commercial capital.Therefore,it is necessary to start from the five dimensions of security,technology,law,ethics and value,so as to give full play to the empowering effect of generative AI on the construction of digital government.

关 键 词：生成式人工智能 数字政府 公共服务 

分 类 号：D63[政治法律—政治学]