The AI for African Population Health Unit (AIPHU) will establish an interdisciplinary research and training programme that integrates genomics and other omics, biomedical imaging, and artificial intelligence to address major non‑communicable and infectious diseases affecting South Africa and the wider African continent.

The unit will develop African‑relevant, ethically grounded AI/ML tools for precision medicine and multimodal diagnostics that account for the unique genetic, clinical and socio‑environmental contexts of African populations.

Its research is structured around three core pillars: AI‑driven multi‑omics precision medicine for biomarker discovery and risk stratification; multimodal AI for biomedical imaging to improve diagnostic accuracy and efficiency in settings with limited specialist expertise; and TorchAfrica, an open‑source PyTorch framework providing pretrained deep‑learning models derived from African biomedical data to improve transfer learning and model performance.

Alongside innovation, AIPHU emphasizes capacity building, sustainability and open science, leveraging established continental training programmes to develop a skilled African AI and biomedical data‑science workforce and to ensure broad access to tools, data and knowledge.