Trust in Chatbot-Generated Health Information: A User Perspective
摘要
With the growing use of AI-powered chatbots, this study investigates user trust in health information retrieved via ChatGPT compared to traditional search engines such as Google. Using a qualitative methodology, 23 undergraduate and graduate students with prior ChatGPT experience completed health-related tasks using both tools, followed by a satisfaction and trust questionnaire and semi-structured interviews. Quantitative findings showed no significant difference in user satisfaction between ChatGPT and Google; however, trust in ChatGPT’s health responses was significantly higher. Qualitative analysis revealed that while users valued ChatGPT’s speed, structure, and ease of interaction, their trust remained cautious and conditional—especially among more educated users—due to concerns over source transparency and accuracy. Many emphasized the need for verification through experts or alternative sources. These findings highlight ChatGPT’s role as a supplementary tool in health information seeking and underscore the importance of enhancing source attribution and digital health literacy to support informed decision-making by users.