African economies risk being left behind by an AI revolution built on data that does not adequately represent the continent, according to the Gates Foundation’s 2026 Goalkeepers report.
More than 90% of the data used to train early large language models came from English-language sources.
The Gates Foundation is committing at least $1 billion over the next two years to try to close that gap, with funding focused on making AI work across more languages and local contexts, as well as in health, education and agriculture.
For Africa, that means building tools around local data and conditions — and ensuring governments, workers and developers have a role in shaping them, the foundation said.
The Gates Foundation provides financial support for Semafor’s Next 3 Billion series.