Health AI Must Be Built Around Africa's Data
Franck Tchafa urged African institutions to develop health AI around local problems and local data rather than systems trained elsewhere.
Updated on Nov. 4, 2026 in Artificial Intelligence
Live Poll
Would you trust an AI tool to help your doctor read your medical scans?
Health technology expert Franck Tchafa told a forum organized by the African Healthcare Association that African institutions should build health AI around local problems and local data. He warned that AI systems trained mostly on other populations may misread African patients.
Why it matters
Africa faces a projected shortfall of 11.8 million health workers by 2030, and AI tools could help with image analysis and triage. However, tools that perform poorly on African patients could worsen existing gaps in care.
The projection of an 11.8 million health worker shortfall by 2030 is a forecast presented at the forum, not a measured current deficit. Africa's population is expected to rise from roughly 1.5 billion to 2.5 billion by 2050, but the sources do not detail how AI adoption rates compare across the continent.
The players
Franck Tchafa
Tchafa is a health technology expert who spoke on the use of AI in African healthcare.
African Healthcare Association
The African Healthcare Association organized the forum where Tchafa spoke and will hold its inaugural conference in Ghana.
The details
Tchafa said barriers to AI adoption include fragmented and paper-based health records, inconsistent data standards, unreliable connectivity and electricity, high costs, a shortage of local AI specialists, and regulatory uncertainty. He urged African countries to leapfrog as they did with mobile money and off-grid solar, and raised concerns about patient privacy, informed consent, algorithmic bias, and over-reliance on automated systems.
Timeline
2030: Projected shortfall of 11.8 million health workers.
2050: Africa's population projected to reach 2.5 billion.
18-19 December 2026: African Healthcare Association holds its inaugural conference in Ghana.
The Tech Race
Tchafa's call to leapfrog legacy health systems echoes the path mobile money took across Africa, skipping traditional banking infrastructure. Whether health AI follows that trajectory depends on the data and connectivity barriers he identified.
For patients, AI tools that were not trained on local populations may misread test results such as retinal images, affecting diagnosis accuracy. Clinics adopting these tools will need to address privacy and informed consent before patient data is used.
The takeaway
Health AI built on local data and local problems is more likely to serve patients accurately than systems imported from elsewhere. Institutions should map their clinical workflows and data quality before adopting any tool.
Further reading
For more on the tools reshaping clinical work, see Artificial Intelligence coverage.
Live Poll
Would you trust an AI tool to help your doctor read your medical scans?
