Conversational Agent-Based CBT Intervention for Anxiety and Depression in Medical Students: A Pilot Randomized Feasibility Study

Background: Medical students experience substantially higher rates of anxiety and depressive symptoms than similarly aged members of the general population. Academic, financial, and social pressures, together with concerns about stigma, may discourage timely engagement with conventional mental health services. Artificial intelligence (AI)-enabled conversational agents delivering cognitive behavioural therapy (CBT) may provide an accessible and scalable form of psychological support, although evidence among international medical students remains limited. Objective: This study evaluated the feasibility and preliminary clinical effectiveness of an AI-supported conversational agent delivering CBT-based interventions for reducing symptoms of anxiety and depression in international medical students. Methods: A two-week pilot randomized feasibility trial was conducted among international medical students with clinically relevant symptoms of anxiety and/or depression. Participants were randomized to either an AI conversational CBT intervention or a psychoeducational self-help control. Feasibility and changes in anxiety and depressive symptoms were assessed following the intervention. Results: Of 228 screened students, 138 were randomized, with 74.6% retained at two weeks. Study completion was substantially higher in the intervention group. ANOVA demonstrated a significant effect of the intervention on post-intervention depressive symptoms, with participants receiving the conversational agent reporting significantly lower scores than controls. Anxiety symptoms also showed a significant intervention effect, with moderate-to-large treatment effects observed after two weeks. Conclusions: AI-supported conversational CBT appears feasible and acceptable for international medical students and may reduce short-term symptoms of anxiety and depression. The significant ANOVA findings provide preliminary evidence of clinical effectiveness. Larger multicentre randomized controlled trials with longer follow-up are warranted to confirm these findings and determine the potential role of AI-based interventions within university mental health services.