Abstract

In recent years, artificial intelligence-based chatbots have become increasingly present in adolescents’ daily lives, used for various purposes, including mental health support. The aim of this study is to review recent literature (2022-2026) on the use of AI chatbots for psychological support among adolescents. Evidence regarding effectiveness, as well as associated risks and clinical and ethical implications, is examined. Findings from recent studies suggest that chatbot-based interventions may lead to modest reductions in psychological distress. However, generative AI systems show considerable variability in response quality and raise concerns related to privacy, misinformation, limited crisis-detection capabilities, social isolation, and potential dependency. This review discusses the main theoretical frameworks, evaluates the available empirical evidence, and analyzes the ethical and clinical challenges in the near future. Future research should focus on longitudinal study designs and the potential clinical validation of hybrid systems with therapist supervision, in order to assess these tools as complementary supports in adolescent mental health care.