Abstract:

With the in-depth advancement of digital government construction and the rapid development of artificial intelligence(AI) technology, especially following the launch of domestic open-source large models, government departments at all levels are actively promoting the deployment and application of AI large models. Large models possess the capability to drive innovations in government operation modes and serve as key tools to improve governance system and enhance governance capacity. By clarifying the classification of government large models and constructing a reference model for their application, this paper illustrates the supporting role of large models in government business work through three cases such as intelligent document approval, intelligent official document generation, and online transaction supervision. In response to the shortcomings existing in the application of government large models, including aspects of planning and design, credibility of generated content, quality of model training data, and data security risks, this paper proposes implementation suggestions for advancing the application of government large models, focusing on key issues such as the construction and deployment of government large models, the development of trustworthy models, the supply of data elements, the opening of application scenarios, security and risk prevention, as well as construction and operation.

Key words: artificial intelligence, large language models for digital government, application