Generative AI Use for Mental Health Support: Trends, Correlates, and Outcomes Among Canadian Students

Abstract General-purpose generative AI (GenAI) chatbots are increasingly used for mental health support, yet population-based evidence on who uses them, why, and how remains limited. We estimated prevalence, patterns, correlates and impact of GenAI use for mental health using repeated cross-sectional data from 1,067 students in the Canadian arm of the WHO World Mental Health International College Student (WMH-ICS) survey. Correlates were assessed using modified Poisson regression, with an elastic-net machine learning model employed as sensitivity analysis. Past-year GenAI use for mental health or emotional support was 25.2% (95% CI 22.7–27.9) and lifetime prevalence was 30.6%. Students mostly used GenAI for mental health information, stress management, and companionship. Use was more prevalent among Asian, Black, and Hispanic students, and those with greater clinical burden, recent adverse life experiences, and prior digital help-seeking. It was less prevalent among LGB students, partnered students, and those with greater social support. Nearly three-quarters of users (74.2%) reported positive impact on their mental health and emotional needs. Non-use reflected preference for human interaction, distrust, privacy concerns, and principled objections to AI, with 83% of never-users indicating no future use intention. The growing use of GenAI for mental health purposes underscores the need for rigorous evaluation and appropriate governance.