Learning from Nigeria’s public-private pathways to generate inclusive healthcare data

Artificial Intelligence (AI) is transforming healthcare, with applications spanning disease diagnosis, patient monitoring, personalized treatments, and drug discovery 1 , 2 . There is growing consensus among healthcare providers, policymakers, and patients that AI-enabled digital health is reshaping care delivery by accelerating diagnoses, improving outcomes, reducing costs, and expanding access to remote regions. Yet, despite these advances, a persistent digital divide threatens to widen health inequities. The datasets powering AI systems remain predominantly sourced from white, Western populations, leaving racial and ethnic minorities and entire regions such as Africa underrepresented 3 . This lack of data diversity produces AI models that are less accurate, less generalizable, and potentially biased, exacerbating disparities in healthcare outcomes 4 . In addition, African Americans are less likely to participate in health research due to longstanding mistrust, as seen in HIV, cancer prevention, and occupational health studies 5 , 6 . The consequence is a problematic cycle with underrepresentation in data leading to underperformance in care delivery 7 .