Conversational AI in Pediatric Mental Health: A Scoping Review of Applications, Evidence, and Special Considerations

Background/Objectives: Mental health disorders among children and adoles-cents represent a significant global health challenge, with approximately 50% of condi-tions emerging before age 14. Despite substantial investment in services, persistent barriers including provider shortages, stigma, and accessibility issues limit effective care delivery. This scoping review examines the emerging application of conversational ar-tificial intelligence (AI) in pediatric mental health contexts, mapping the current evidence base, identifying therapeutic mechanisms, and exploring unique developmental con-siderations required for implementation. Methods: Following established scoping review methodology, we searched multiple electronic databases (PubMed/MEDLINE, PsycINFO, ACM Digital Library, IEEE Xplore, and Scopus) for literature published between January 2010 and February 2025 that addressed conversational AI applications relevant to pedi-atric mental health. We employed a narrative synthesis approach with thematic analysis to organize findings across technological approaches, therapeutic applications, devel-opmental considerations, implementation contexts, and ethical frameworks. Results: The review identified promising applications for conversational AI in pediatric mental health, particularly for common conditions like anxiety and depression, psychoeducation, skills practice, and bridging to traditional care. However, most robust empirical research has focused on adult populations, with pediatric applications only beginning to receive dedicated investigation. Key therapeutic mechanisms identified include reduced barriers to self-disclosure, cognitive change, emotional validation, and behavioral activation. Developmental considerations emerged as fundamental challenges, necessitating age-appropriate adaptations across cognitive, emotional, linguistic, and ethical dimen-sions rather than simple modifications of adult-oriented systems. Conclusions: Conver-sational AI has potential to address significant unmet needs in pediatric mental health as a complement to, rather than replacement for, human-delivered care. Future research should prioritize developmental validation, longitudinal outcomes, implementation science, safety monitoring, and equity-focused design. Interdisciplinary collaboration involving children and families is essential to ensure these technologies effectively ad-dress the unique mental health needs of young people while mitigating potential risks.