Public Servants' Professional Development and AI Competencies - the Case of Statutory Training for Promotion in Taiwan

• In Event: AI and the Public Workforce – Skills, Concerns, and Services

Abstract

Since the emerging application of ChatGPT as well as more generative artificial intelligence (AI) tools in late 2022, AI has become not only the fancy buzz word, but also true influential productivity tools for public servants. However, there remain few empirical studies investigating (1) the extent public servants influenced by AI (2) the actual and expected AI use cases, and (3) the institutional adaption of AI adoption in the public sector. Moreover, how can the current training and developing scheme of public administration transform and integrate with the broadening AI use scenarios?

The current study, with the delegation and assistance from Civil Service Protection and Training Commission (CSTPC) in Taiwan, conducted a survey involving over 3,000 public servants who had just taken the statutory training for promotion program for their AI use intention and behavior in their personal purposes and public tasks. Also, we developed a five-module framework of AI competencies for public servants training, aiming for their career development suitable for different public service levels. Meanwhile, the senior IT managers in the public sector were surveyed to rank the five-module framework of AI competencies aforementioned to ensure its completeness and feasibility for professional development and AI training programs.

This empirical results show that, first, over 70% of the surveyed public servants had used any AI-based online service or software before, but only around 25% used for their tasks in the public sector. Secondly, the most used AI functions include producing transcription, summarizing and translating, followed by drafting press release or script. The most challenging and expected AI uses in public sector are document processing, resource integration and data analysis. Also, generally the public servants in any rank would like more AI training programs. Finally, all public servants report positive attitude toward future development of AI in general, and the positive attitude ascending by their rank, which means the higher the rank, the most positive attitude toward AI, both for the society or for the government.

Overall, the study contributes to the AI adoption in public administration in three aspects. First, based on the first hand survey data, public servants’ intention and actual behavior of AI are vividly presented and categorized, which is essential for the future AI adoption AI. Secondly, the proposed five-module AI literacy empirically evaluated can be generalized not only for the public sector in Taiwan, but also in different countries. Third, the five modules empirically confirmed by the senior IT managers strengthen the validity and usefulness of the AI training programs, and also result in customized AI course designs for public servants worldwide.

Authors

• Hao-En Kao, YuanZe University
  Presenting Author

• Naiyi Hsiao
  Non-Presenting Co-Author