Artificial intelligence for precision malaria control: transforming surveillance, prediction, and intervention strategies

Abstract

Malaria elimination has stalled globally despite decades of investment, with traditional surveillance constrained by retrospective reporting and limited capacity to integrate high-dimensional, non-linear data. In this perspective, we argue that artificial intelligence (AI) encompassing machine learning and deep learning can help shift malaria control from reactive reporting toward predictive, precision public health, while cautioning that it is one enabler among many rather than a stand-alone solution. We examine AI across three interconnected domains: surveillance (understood broadly as case detection, entomological and intervention-coverage monitoring, and the data-to-response loop), prediction (outbreak forecasting and spatial risk mapping), and control (resource stratification and intervention optimisation). Reported AI diagnostics can exceed 90% accuracy, and forecasting models offer useful lead times by integrating climatic, entomological and epidemiological data. However, realising this potential depends on resolving data-quality limitations, algorithmic bias, weak digital infrastructure, and absent governance frameworks, and on locally adapted rather than uniformly generalised models. We contend that demand-driven integration into national strategies, local capacity building, and prospective trials not algorithmic novelty will determine whether AI meaningfully accelerates progress toward elimination.