Patient trust in medical AI: Conditions under which chatbots are preferred over human providers

BACKGROUND: Artificial intelligence (AI) tools, including medical chatbots, are increasingly integrated into health care delivery. While prior research emphasizes diagnostic accuracy and usability, less is known about the psychological and contextual factors that shape patients' trust and preferences for AI versus human providers, particularly among patients who have experienced medical stigma. PURPOSE: This study investigates when and why patients may prefer medical chatbots over human providers, focusing on the roles of prior experiences of stigma, comfort with medical AI, and the contextual tradeoff of wait time. METHODOLOGY/APPROACH: Two scenario-based experiments examined trust in a chatbot versus a human provider (N=300) and choice between a chatbot and a human provider depending on wait time (N=450). Measures included trust in the provider (perceived accuracy and bias), past experiences of medical stigma, and comfort with medical AI. RESULTS: In general, participants preferred human providers. However, participants who reported high comfort with medical AI and those with prior experiences of stigma (e.g., related to weight, mental health, or gender identity) expressed greater trust in the chatbot. Moderation analysis confirmed significant interaction effects between provider type, comfort with medical AI, and stigma history. CONCLUSIONS: Preferences for AI-driven care are conditional. Chatbots may offer psychological safety and perceived impartiality for patients with stigmatized identities. PRACTICE IMPLICATIONS: Providers are often challenged with managing high caseloads and allocating appointments. Integrating chatbots in care pathways could improve access to care and enhance trust among marginalized populations, especially for low-stakes concerns, offering cost-effective, patient-centered alternatives in selected contexts.