LARGE LANGUAGE MODEL–BASED CHATBOTS IN MENTAL HEALTH SUPPORT: A NARRATIVE REVIEW OF OPPORTUNITIES, RISKS, AND ETHICAL CHALLENGES
Abstract
Background: Mental disorders represent a major global public health challenge. Mental health care is affected by limited access to services, shortages of mental health professionals, stigma, and growing healthcare costs. In response to these challenges, artificial intelligence-based technologies, including chatbots and large language models, are increasingly being explored as tools to support mental health care. Aim: The aim of this narrative review was to examine the potential benefits, risks, and ethical challenges associated with the use of artificial intelligence chatbots and large language model-based systems in mental health care. Particular attention was given to their clinical relevance, public health implications, and responsible implementation. Methods: A narrative review of the literature was conducted using PubMed, Scopus, Web of Science, and Google Scholar. Publications written in English and published between January 2013 and March 2026 were considered. Thematic synthesis focused on technological applications, clinical utility, system-level implications, risks, and ethical and regulatory challenges. Results: Current evidence suggests that AI-based chatbots may support psychoeducation, emotional support, self-management, symptom monitoring, cognitive behavioral therapy-based exercises, and access to low-intensity mental health support. These tools may be particularly useful in settings with limited access to specialists and for individuals facing barriers related to stigma, cost, or availability of services. However, their use is associated with important risks, including inaccurate or harmful responses, privacy concerns, algorithmic bias, limited transparency, unclear accountability, emotional dependence, and insufficient long-term clinical validation. Conclusion: AI chatbots and large language model-based systems should be considered adjunctive tools rather than replacements for mental health professionals. Their safe and ethical integration into mental health care requires clinical validation, human oversight, transparent governance, privacy protection, bias assessment, crisis protocols, and continuous monitoring.