Artificial Intelligence-Based Digital Interventions for HIV Prevention in Sexual and Gender Minority Populations Globally: A Systematic Review.

Abstract

Abstract: Sexual and gender minority populations experience high HIV incidence, and digital interventions offer prevention opportunities. Artificial intelligence (AI) enables adaptive, personalized support, yet quantitative evidence in sexual and gender minority communities remains limited. This systematic review examined the effectiveness and implementation of AI-powered digital HIV prevention interventions. Five databases were searched from inception to June 2025 for randomized trials, quasi-experimental studies, and feasibility studies using adaptive AI features. Risk of bias was assessed with RoB 2, ROBINS-I, and NIH tools. Findings were synthesized narratively. From 380 records, seven studies (n = 1,729) met the inclusion criteria. Interventions included AI-driven chatbots, adherence monitoring systems, and a simulation game targeting preexposure prophylaxis adherence, HIV testing, and sexual risk reduction. Chatbots showed the consistent behavioral effects, particularly in increasing HIV self-testing and preexposure prophylaxis uptake. Across studies, feasibility, engagement, and usability were high. AI-based digital interventions show promise, but rigorous trials are needed to strengthen the evidence.