Review Article

Artificial Intelligence and Predictive Analytics for HIV Surveillance: From Retrospective Monitoring to Anticipatory Outbreak Response

Abstract

Despite major advances in HIV testing, antiretroviral therapy (ART), pre-exposure prophylaxis (PrEP), and surveillance, ongoing transmission and persistent gaps across the HIV care continuum remain important public-health challenges. Most surveillance systems, however, remain predominantly retrospective: they describe diagnoses, transmission clusters, care disengagement, and viral non-suppression after these patterns have already emerged. Artificial intelligence (AI) and predictive analytics offer the possibility of shifting surveillance toward earlier identification of emerging risks, but improved predictive performance does not necessarily translate into effective prevention. This critical narrative review examines the evolving role of AI in HIV surveillance, integrating evidence across routine surveillance and care-continuum data, molecular and genomic surveillance, clinical and electronic health records, and behavioral, social, geographic, and mobility information. Particular attention is given to data latency, prediction horizon and lead time, concept drift, transportability, multimodal integration, the prediction-to-response gap, and ethical and equity concerns. The review argues that the central challenge is no longer simply whether HIV-related outcomes can be predicted, but whether predictions provide sufficiently early, validated, interpretable, equitable, and actionable intelligence to change public-health decisions and outcomes. To address this gap, I propose an AI-enabled anticipatory HIV surveillance framework organized around  BSERVE → INTEGRATE → DETECT → PREDICT → VALIDATE → PRIORITIZE → RESPOND → EVALUATE → LEARN. Future progress should therefore be evaluated by the ability of predictive surveillance to generate meaningful lead time and enable interventions that improve HIV prevention and care.

Authors affiliation:

Ndubisi Ubaka Edebeatu (ORCID): Department of Medicine, Texila American University, Guyana.

How To Cite: N.U. Edebeatu. Artificial Intelligence and Predictive Analytics for HIV Surveillance: From Retrospective Monitoring to Anticipatory Outbreak Response. Infectious, 3(2):27–42, 2026. https://doi.org/10.58612/infectious324