Do College Students Prefer AI Health Coaches Over Human Health Coaches? – A Mix-Method Exploratory Study

Background Rising rates of chronic disease, mental health concerns, and lifestyle-related challenges among college students highlight the need for scalable approaches to support student well-being.Purpose This study developed and tested the feasibility of an artificial intelligence (AI) virtual health coach and compared it to human coaches among college students.Methods Thirty-one college students completed six coaching sessions with seven trained human coaches and two consecutive sessions with an AI coach. The AI agent was developed using prompt engineering on a commercial large language model platform. After each session, students completed quantitative evaluations. Semi-structured interviews were conducted.Results Students rated human coaching sessions significantly higher than AI sessions across all evaluated dimensions. Rater evaluations indicated comparable performance between AI and human coaches, but human coaches demonstrated significantly greater empathy and helpfulness. Qualitative findings revealed that human coaches were valued for care, connection, accountability, and responsiveness, whereas AI coaches were perceived as nonjudgmental, accessible, and highly structured, with strengths in goal clarification and information delivery.Discussion Whereas human coaches were preferred overall, AI health coaching demonstrated feasibility and potential value as a complementary, scalable support tool.Transition to Health Education Practice Practitioners should proactively seek training in AI tools, prompt design, and safety-tuning so they can effectively and responsibly deploy AI as a supplemental tool for the health coaching practice.