Utilizing Artificial Intelligence in Personal Health Contexts: Investigating Information Seeking Motivations, Information Processing Behaviors, and Behavioral Responses During and After Interaction with GenAI

Following the advent of generative artificial intelligence (GenAI), seeking health information on GenAI has become a more common and flexible approach. However, laypeople's health information seeking and processing (HISP) motivations and behavior on GenAI platforms remain underexplored, particularly given the current shift from search-based activities to more interactive and conversational processes. This study addressed this gap by exploring laypeople's HISP motivations, behaviors, and behavioral responses during and after interaction with GenAI. Using the risk information seeking and processing model as the theoretical framework, the HISP process was explored from a risk reduction perspective to understand how laypeople interact with GenAI to address their health concerns. Think-aloud protocol and retrospective interviews were employed to capture participants' real-time and reflective HISP interactions. Participants were mainly young individuals in Hong Kong. Through systematic thematic analysis using both inductive and deductive coding, we found that participants' GenAI-based health information seeking was shaped by multifaceted factors, including relevant health experiences, perceived health characteristics, affective responses toward health risks, and personal factors such as relevant channel beliefs. Trust played a critical role in shaping how participants processed the provided information and was associated with indicative heuristic and systematic processing patterns, given the information characteristic (medical vs non-medical) and perceived acceptance risks. Emotion sufficiency, information overload, and information affordance were identified as emerging variables and discussed in the GenAI-mediated HISP context. Finally, participants may choose to follow, reject, or integrate the GenAI-provided health advice into future health consultations, considering information characteristics and implementation difficulty.