Bridge or barrier? The impact of GAI chatbot-based pre-consultations on patients' subsequent trust in physicians

Purpose Generative artificial intelligence (GAI) chatbots are increasingly used as pre-consultation tools in healthcare. However, while prior research has primarily explored patients' adoption of GAI chatbots or their effects on health outcomes, the impact of GAI chatbot-based pre-consultations on patients' subsequent trust in physicians remains underexplored. This study examines how perceived information quality and privacy concerns about GAI chatbots during pre-consultations shape patients' perceived usefulness of GAI chatbots and, in turn, influence their subsequent trust in physicians. We also investigate the moderating roles of patients' speciesism and health literacy. Design/methodology/approach Drawing on trust theory, this study develops a research model. It analyzes survey data from 356 individuals in China who used GAI chatbots before consulting a physician using partial least squares structural equation modeling (PLS-SEM). Findings The results show that GAI chatbot-based pre-consultations have a double-edged effect driven by a benefit-risk mechanism: high-quality information enhances perceived usefulness, which in turn strengthens physician trust, whereas privacy concerns undermine subsequent trust in physicians by reducing perceived usefulness. Moreover, the positive relationship between perceived usefulness and subsequent trust in physicians is stronger among patients with lower levels of speciesism and health literacy. Originality/value By identifying a novel pathway through which patient–AI interactions during pre-consultations shape subsequent physician–patient trust, this study highlights the double-edged nature of the use of GAI chatbots, extends prior research on GAI chatbots and physician–patient trust, and offers actionable implications for the design of GAI-enabled pre-consultation tools.