AI in Mental Health: Literature Review

The increasing prevalence of mental health disorders such as depression, anxiety, and stress has become a significant global challenge, while access to professional mental health services remains limited in many regions. Recent advances in artificial intelligence, particularly large language models (LLMs), offer new opportunities to provide scalable, accessible, and efficient mental health support through psychological data analysis, conversational systems, and clinical decision assistance. This study aims to systematically examine the development of research on the application of LLMs in mental health support and to identify the main applications, benefits, and ethical challenges associated with their implementation. The research employs a systematic literature review and bibliometric analysis approach following the PRISMA guidelines. Data were collected from the Scopus database using keywords related to artificial intelligence, large language models, and mental health. Articles that met the inclusion criteria were published between 2020 and 2026 and were analysed based on research objectives, methodologies, main variables, and research findings. The results indicate that LLMs have significant potential to support early detection of mental health disorders, predict depression risk, analyse digital behavioural data, and provide psychological assistance through AI chatbots and digital advisory systems. Furthermore, these technologies contribute to improving mental health literacy and supporting clinical decision-making processes. However, several studies highlight challenges related to information accuracy, algorithmic bias, data privacy, and the need for professional supervision when implementing AI-based mental health systems. Overall, the findings suggest that LLMs can play an important role in the development of digital mental health services, but their implementation must be accompanied by strong ethical frameworks, rigorous clinical validation, and collaboration between technology developers and mental health professionals to ensure safe, accurate, and responsible use.