The Science That Skips Africa: a continent of 1.4 billion produces 1% of the world's research — and is missing from its data

There is a researcher in a university lab somewhere in Africa who is as sharp as any in the world, working on a problem the world urgently needs solved — a local crop disease, a regional health threat, a question only her context can answer. She works with a fraction of the funding, a sliver of the equipment, intermittent power, and almost none of the data infrastructure her peers abroad take for granted. Her findings, when she manages to publish them, join a vanishingly small share of the world's science. And the data she could generate — about African people, languages, soils, diseases — is largely absent from the datasets the rest of the world builds its knowledge on. That absence has consequences for everyone.

A continent missing from the world's research

The investment gap is stark. African countries spend on average about 0.4% of GDP on research and development — roughly a quarter of the global average of 1.7%. Sub-Saharan Africa has only around 88 researchers per million people, a tiny fraction of the rate in Europe or North America. And the continent produces under 1% of the world's research output. For 17% of humanity and the fastest-growing population on Earth, that is not a knowledge economy — it is a knowledge deficit, and it compounds: fewer researchers means less research, less research means less funding, less funding means fewer researchers.

And missing from its data

The deeper problem is representation. Modern science — and increasingly, modern AI — is built on data, and African data is barely in it. As recently measured, only about 3% of the human genomic data used in global studies comes from Africans, even though Africa holds the greatest human genetic diversity on the planet. The same gap runs through medical, linguistic, agricultural and climate datasets. When drugs, diagnostics and AI models are trained on data that excludes African populations, they work less well for Africans — sometimes dangerously so. A science that skips Africa doesn't just shortchange the continent; it builds a less accurate picture of humanity for everyone.

AI is the great equalizer for the under-resourced lab

Here is what has changed. For most of history, doing world-class research required world-class resources — armies of assistants, expensive software, proximity to the centres of knowledge. AI collapses much of that. A single researcher can now have a tireless assistant that reads the literature, drafts and debugs analysis code, structures messy data, and surfaces connections across thousands of papers. The under-resourced lab is precisely where that leverage matters most — and, paired with tools to gather and build African datasets, it lets the continent's researchers both do far more and finally put African data, languages and contexts into the global record.

The fix for a research gap this wide was always going to be leverage, not just money. AI lets a handful of African researchers do the work of many — and the data they build is how the world's science stops skipping a continent.

Research that includes Africa: the toolkit

Amplifying African research takes a connected kit — automate the grind, access the models and compute, build datasets in African languages, gather data from communities, and train the next generation. Here's the fit.

The prize

A continent that under-invests in research and is missing from the world's data is leaving its hardest problems for others to solve — others who lack the context, the data, and the stake to solve them well. Flip it, and the prize is enormous: African researchers, amplified by AI and armed with African data, working on African problems — drought-resistant crops, local disease genomics, climate adaptation, language technology — and contributing discoveries the whole world needs. The talent is already here. Give it leverage and put its data on the map, and the science that skips Africa becomes science the continent leads. That is the prize worth funding.