Artificial intelligence-driven interventions to improve mental health outcomes: an umbrella review of meta-analyses

This umbrella review aims to synthesize high-level evidence on the efficacy and engagement of artificial intelligence (AI)-driven interventions for improving mental health outcomes.We will systematically search PubMed, Embase, PsycINFO, and the Cochrane Library for meta-analyses published through March 2026. Methodological quality will be assessed using the AMSTAR 2 tool, and evidence certainty will be graded via the GRADE approach. A citation matrix will be constructed to calculate the Corrected Covered Area (CCA) to address primary study overlap.The findings will provide a comprehensive evidence map of AI’s role in mental health care, identifying which psychiatric conditions benefit most from AI-driven support.