AFRICA’S STAKE IN THE AI REVOLUTION: WHY BUILDING LOCAL CAPACITY MATTERS FOR EVERYONE
African data is already training the world’s most powerful AI models. The intelligence those models generate flows back to the continent as a product Africans pay for, built on realities they never validated and optimized for contexts that were never theirs. That, in essence, is the warning Kate Kallot has been sounding from her position at the intersection of technology, policy, and African development.
Kallot, a Kenyan techpreneur and founder of Amini, an African data infrastructure company, frames the challenge in terms that echo a much longer historical pattern: extraction without ownership, raw materials without the value that comes from processing them. “African data trains the models, the intelligence flows back as a product Africans pay for, built on realities Africans never validated; optimized for contexts that were never ours,” she explains. The concern is not merely about technology access or innovation hubs. It is about whether Africa will repeat a centuries-old cycle in which the continent supplies essential resources while others build the industries and capture the wealth.
The numbers underscore the disparity. Africa represents roughly 18% to 19% of the world’s population but accounts for less than 1% of global compute capacity. The general-purpose AI models reshaping industries worldwide are trained primarily on internet data, a resource that remains inaccessible to millions of Africans who lack connectivity. This gap creates a dual vulnerability: the continent’s data and insights are being used to train systems that do not reflect African realities, while Africans themselves have little say in how those systems are built or deployed.
Kallot’s work through Amini focuses on building sovereign data and compute systems designed specifically for Africa and the Global South. Her advocacy centers on a straightforward proposition: if the continent does not develop its own AI infrastructure and capacity, it will face consequences that extend far beyond technology. “If Africa does not build its own AI future, the AI divide will become the social divide,” she says. “And we will have missed the single greatest economic opportunity this continent has ever had, while repeating the most painful pattern in our history: brilliant people; abundant resources; value extracted elsewhere.”
The geopolitical dimension adds another layer of urgency. Kallot points out that AI development is inseparable from questions of national power and sovereignty. “We have entered an era where you cannot decouple AI from geopolitics,” she notes. “And, if your government’s digital infrastructure runs on platforms owned by a foreign power, that foreign power holds the kill switch on your economy.” That framing transforms the AI conversation from a purely technical or economic one into a matter of national security and self-determination.
The parallel Kallot draws to earlier patterns of resource extraction is deliberate. Africa has historically supplied raw materials: cacao, cobalt. In each case, the continent provided the source material while other regions built the industries, created the jobs, and accumulated the profits. Data, in Kallot’s analysis, represents the latest iteration of that same dynamic. If Africa does not control how its data is used and who builds the systems trained on it, the pattern will simply continue in a new technological form.
By contrast, Kallot’s own trajectory suggests the argument is gaining traction well beyond the continent. She was named one of TIME’s 100 Most Influential People in AI in 2023 and received the One Young World Entrepreneur of the Year award in 2024. She currently serves as Vice Chair of the International Chamber of Commerce’s Global Environmental and Energy Commission and sits on EY’s Global AI Advisory Council.
The challenge ahead is not simply technological but structural. Building the infrastructure, talent, and institutional capacity for Africa to develop its own AI systems requires sustained investment, policy support, and regional cooperation. Without it, the continent risks becoming a data source for systems built elsewhere, trained on African realities but optimized for other contexts, and sold back to African users at a premium. Whether the policy commitments and investment needed to break that cycle will materialize, and how quickly, remains the open question that will define the continent’s position in the AI era.