Kuta B, Novak L, Zidkova R, et al. Effectiveness of a Fully Automated Mobile Therapeutic Versus a General Chatbot in Reducing Depression and Anxiety and Improving Well-Being: Feasibility Randomized Controlled Trial. JMIR Ment Health. 2026;13:e82642. doi:10.2196/82642
This pilot study examined whether a mental health chatbot app (ChatMind) tailored to provide therapy could improve mental health outcomes more effectively than a general-purpose chatbot or no intervention. ChatMind was rooted in Solution-Focused Brief Therapy (SFBT), a counseling approach that uses structured, goal-oriented conversations to help people identify solutions and make positive changes. Participants (n=147; convenience sample) were randomly assigned (1:1:1) to one of three groups for the three-week intervention period. Most participants were women, and none were actively seeking mental health treatment. One group used the AI therapy chatbot through the ChatMind app, another used ChatGPT for similar conversations, and a third group received no intervention. Participants completed questionnaires before and after the study to measure symptoms of depression and anxiety, as well as overall well-being. Each week, participants were asked to complete three chatbot sessions, but most did not complete all the required sessions. Results showed that both chatbot groups experienced significant reductions in symptoms of depression compared with the control group. Both groups also showed small improvements in anxiety and well-being, although these changes were not statistically significant. ChatMind did not significantly outperform ChatGPT on any measured outcome, although it showed slightly larger improvements in specific measures of depression and anxiety. This study highlights the potential of generative artificial intelligence as a tool for supporting mental health. However, participant engagement was lower than expected, especially with ChatMind. These findings suggest that AI-based mental health tools may be helpful, but more research is needed to determine which design features and therapeutic approaches are most effective.