AI-based chatbots and public value creation in digital government services

Purpose – Governments increasingly deploy artificial intelligence (AI)-based chatbots to improve accessibility, efficiency and consistency in citizen services. However, empirical evidence on how such systems contribute to public value and service process transformation remains limited. This study aims to examine how a government-operated chatbot shapes citizen interactions and public value through emotional, thematic and behavioral dynamics. Design/methodology/approach – The study analyzes 1, 846 real-world inquiries submitted to a Ministry of Labor chatbot. Sentiment analysis, computational content classification and longitudinal trend modeling were used to assess emotional tone, topic concentration, usage patterns and repeat-contact behavior. Findings – Citizen inquiries were predominantly neutral or negative in tone, while chatbot responses remained largely neutral, indicating emotional stabilization rather than escalation. Employment security topics generated higher emotional intensity, although sentiment differences across topics were not statistically significant. Usage fluctuated over time, and repeated inquiries increased moderately, suggesting both informational friction and growing institutional reliance. Practical implications – Effective public-sector chatbot deployment requires process-oriented governance, topic-sensitive response design, explainability mechanisms and continuous performance monitoring to align operational efficiency with citizen experience. Originality/value – The study advances understanding of AI-enabled service transformation by operationalizing public value through interaction-level indicators and demonstrating how chatbot-mediated processes influence citizen experience in digital government.