Algorithmic governance in artificial intelligence-driven health systems: A southern African perspective

Abstract

The use of algorithmic systems in public health is expanding rapidly in southern Africa, particularly in areas such as disease prediction, resource allocation and personalised healthcare. While these technologies offer efficiencies, their outputs are shaped by the quality, representativeness and governance of the underlying data. Tools such as polygenic risk scores and pathogen genomics often underperform in black African populations owing to under‐representation in genomic datasets, limiting clinical accuracy and access to precision health benefits. In low‐ and middle‐income countries, these disparities are amplified by fragmented health data systems, limited digital infrastructure, and weak regulatory oversight. Such conditions undermine the ability to detect and correct algorithmic bias, increasing the risk that digital health tools perpetuate inequities affecting rural, low‐income and marginalised populations. In this context, algorithmic systems may unintentionally reinforce existing health disparities. This commentary draws on the concept of biopolitics to highlight how algorithmic tools classify and manage populations through data‐driven processes. It emphasises the need for greater transparency, inclusion and accountability in the development and deployment of these technologies. Aligning health technology assessment processes with equity‐driven metrics and locally governed standards is essential to ensure that algorithmic governance supports, rather than undermines, equitable public health outcomes in the region.