Beyond Bureaucrats: Exploring Passive and Symbolic Representation in AI-Powered Public Administration

• In Event: Insights on Artificial Intelligence in Government

Abstract

Can the findings on passive and symbolic representation in representative bureaucracy be extended to AI? This question has become increasingly relevant as governments adopt AI in public administration (Miller & Keiser, 2021). The 2022 UN E-Government Survey reveals that over 69 countries now use AI chatbots to engage with citizens. Unlike traditional citizen-bureaucrat interactions, where shared demographics or social identities often shape perceptions, AI chatbots lack identity markers such as race, gender, or age. This raises critical questions about whether citizens perceive in-group or out-group distinctions with AI, and how these perceptions affect their willingness to collaborate in AI-driven public services (Gaozhao et al., 2024).

This research employs an experimental design to address these questions by manipulating three factors: (1) participants’ awareness of interacting with AI, (2) the AI’s visual or perceived identity, and (3) the information and outcomes produced by the interaction. The study evaluates public perceptions of the AI agent, the interaction itself, and the associated government agency.

The study draws on anthropomorphism theories, which suggest that humanizing machines enhances relatability. These concepts align with passive and symbolic representation in bureaucracy, where bureaucratic characteristics mirror constituents’ identities. Exploring this intersection offers insights into how AI can foster trust and representation in public services. The findings will bridge theories from public administration and human-machine interaction from computer science, broadening the application of representative bureaucracy to AI contexts. This research has significant theoretical and practical implications, providing actionable insights for designing effective AI systems and supporting government digitization.

References

Gaozhao, D., Wright, J. E., & Gainey, M. K. (2024). Bureaucrat or artificial intelligence: people’s preferences and perceptions of government service. Public Management Review, 26(6), 1498-1525.

Miller, S. M., & Keiser, L. R. (2021). Representative bureaucracy and attitudes toward automated decision making. Journal of Public Administration Research and Theory, 31(1), 150-165.

Author

• Dongfang Gaozhao, University of Dayton
  Presenting Author