生成式人工智能引入数字政府建设的规制路径——基于风险社会的视阈

傅建平

生成式人工智能引入数字政府建设的规制路径——基于风险社会的视阈

引用本文：傅建平.生成式人工智能引入数字政府建设的规制路径——基于风险社会的视阈[J].湖南大学学报社会科学版,2025,(1):124-132

摘要点击次数: 67

全文下载次数: 40

作者	单位
傅建平	（中南大学马克思主义学院，湖南长沙 410083）

中文摘要:数字技术的迭代发展与能力跃迁契合了数字政府转型中的多维需求。生成式人工智能技术能够为数字政府的建设提供在政府决策科学性与前瞻性强化、政府运作效率优化与成本降低、政务服务的亲切度与参与度提升等方面的助力。从风险社会的理论出发，信息时代的风险网络更为复杂、治理更加困难，需要引入三维视角从而综合施策。回归到生成式人工智能引入数字政府建设的场景来看，存在着复杂算法引发的数据安全与信息保密风险、资本力量解构行政权力公共属性的风险、数据处理导致的公民数据权利被侵犯风险。数据分类分级、算法伦理审查、技术问责能够成为应对上述风险的有效规制路径，而针对上述路径，能够以数据多维审视为基点、以行政机关义务为途径、以算法应用监管为场景，提出可操作性更强的细化对策。

中文关键词:生成式人工智能数字政府风险社会法治对策算法治理

The Regulatory Path of Introducing Generative Artificial Intelligence into Digital Government Construction: From the Perspective of Risk Society

Abstract:The iterative development and capability leap of digital technology align with the multidimensional needs in the transformation of digital government. The generative artificial intelligence technology can provide significant support for the construction of digital government by enhancing the scientific and forward-looking nature of government decision-making, optimizing government operational efficiency and reducing cost, and improving the friendliness and participation in government services. However, from the theoretical perspective of risk society, the risk network in the information age is more complex and governance is more challenging, requiring a three-dimensional perspective for comprehensive policy implementation. Returning to the scenario of introducing generative artificial intelligence into digital government construction, there exist risks such as data security and information confidentiality breaches caused by complex algorithms, the risk of capital forces deconstructing the public nature of administrative power, and the risk of citizen data rights being infringed due to data processing. To address these risks, effective regulatory paths include data classification and grading, algorithmic ethics review, and technical accountability. Regarding these paths, we can propose more operational and detailed countermeasures based on multidimensional data review as the foundation, administrative agency obligations as the pathway, and algorithm application supervision as the scenario.

keywords:generative artificial intelligence digital government risk society legal countermeasures algorithm governance

查看全文 查看/发表评论 下载pdf阅读器