Enhancing Public Health Policy Communication Through Government–Citizen Social Media Interactions: The Impact of Replying Agents, Inquiry Tone, and Institutional Trust

This study examines how government responses to citizen inquiries on governmental social media enhance policy communication campaigns for observers. Drawing upon theories of masspersonal communication and artificial intelligence (AI)-mediated communication, it investigates the effects of replying agents, inquiry tones, and institutional trust on observers’ perceptions of the campaigner, campaign message, and communicated policy. A pre-registered four (replying agents: AI-powered bot editor, human editor, AI–human tandem editor, and control) × two (inquiry tones: friendly vs. frustrated) between-subjects online survey experiment was conducted in Hong Kong (N = 1458), focusing on a recent waterpipe smoking control policy. Results revealed that audiences viewed the institution more favorably when friendly inquiries were addressed. AI–human collaborative responses enhanced policy communication effectiveness. Institutional trust moderated the impact of both inquiry tones and replying agents. These findings emphasize that beyond technological factors, nurturing citizens’ confidence in government institutions is crucial for effective AI-mediated government–citizen communication and its role in policy communication.