The Evaluation of Efficacy, Feasibility, and Technical Outcomes of a GPT-4o-based Chatbot Amanda for Relationship Support: A Randomized Controlled Trial

This randomized controlled trial evaluated the efficacy, feasibility, and technical outcomes of Amanda, a GPT-4-based chatbot, in delivering single-session relationship interventions. A total of 258 participants were randomly assigned to engage with either Amanda (n = 130) or a writing task (n = 128) focused on conflict reappraisal. Findings revealed significant improvements in relationship functioning, problem-specific outcomes, and individual well-being in both conditions, supporting the effectiveness of the interventions. Usability and therapeutic skills were rated highly, with participants reporting positive engagement with Amanda. However, technical issues, such as repetitive responses, and the chatbot’s limited ability to identify safety concerns highlight areas for improvement. The results suggest that chatbots like Amanda offer a promising, scalable tool for relationship support, though further research is required to assess long-term effectiveness and address safety concerns in higher-risk populations.