生成式人工智能引入数字政府建设的风险与规制路径——基于TOE理论框架

The Risks and Regulatory Paths of Introducing Generative Artificial Intelligence in Digital Government Construction—Based on the TOE Theory Framework

摘要: 生成式人工智能凭借数据融合、智能生成与自迭代能力，正深刻重塑数字政府的治理逻辑与运行范式，推动政府治理从人工决策加速转型为人工智能决策。然而，在生成式人工智能应用的过程中引发了技术、组织、环境多重风险，这些风险对公民权利保障和政府公信力都构成了挑战。基于TOE (技术–组织–环境)理论分析框架，本文系统剖析生成式人工智能嵌入数字政府过程中存在的风险困境并提出规制路径，有助于推动生成式人工智能与数字政府建设的良性互动，助力提升政府治理效能与治理现代化水平。

Abstract: Generative artificial intelligence, through its capabilities of data integration, intelligent generation, and self-iteration, is profoundly reshaping the governance logic and operational paradigm of digital governments, accelerating the transformation of government governance from manual decision-making to artificial intelligence decision-making. However, during the application of generative artificial intelligence, multiple risks such as technical, organizational, and environmental risks have emerged, posing challenges to the protection of citizens’ rights and the credibility of the government. Based on the TOE (Technology-Organization-Environment) theoretical analysis framework, this paper systematically analyzes the risk dilemmas existing in the process of embedding generative artificial intelligence in digital governments and proposes regulatory paths, which is conducive to promoting the positive interaction between generative artificial intelligence and digital government construction, and helping to enhance the governance efficiency and modernization level of the government.