Teaching decision-making in public administration using AI: A comparative study of student and ChatGPT approaches

The rapid integration of artificial intelligence (AI) into higher education is transforming how decision-making is taught, particularly in fields such as public administration, where complexity, uncertainty, and accountability are central. This study examines how students engage with AI-supported decision-making by comparing human and AI-generated approaches in two educational contexts: public management and business management. Using an experimental classroom design, students were asked to solve a structured decision-making task, apply a selected method, and compare their solutions with recommendations generated by ChatGPT. The study analyzes differences in decision-making methods, outcome alignment, and student satisfaction, complemented by qualitative reflections. The findings reveal significant differences between human and AI approaches to decision-making, while outcome alignment varies independently of methodological similarity. Student satisfaction is strongly associated with agreement between human and AI decisions, suggesting the presence of confirmation effects. Differences between public and business management students are more pronounced in qualitative reflections, particularly regarding the importance of contextual, ethical, and emotional considerations. The results highlight the importance of critically engaging with AI in decision-making education. Rather than replacing human reasoning, AI tools can support reflective learning by enabling comparison, evaluation, and deeper understanding of decision processes. The study contributes to emerging research on AI in public administration education by linking empirical findings to theoretical perspectives on bounded rationality and algorithmic decision-making.