Artificial intelligence could become one of the largest growth engines Africa has ever seen, or it could widen the gap between the continent and the rest of the world. That dual message runs through the World Bank's latest Africa Economic Update, published in early October, which examines how ready Sub-Saharan economies are to harness the technology.
Under an ambitious scenario, the report projects that AI could add up to $1 trillion to Africa's gross domestic product by 2035 and create between 35 million and 40 million digital jobs. The African Development Bank has offered a similar estimate for inclusive adoption. Yet the same document makes clear that the path to that figure is narrow and depends on investments that have barely begun.
Three hubs dominate
The bank found that AI activity in the region is heavily concentrated in Nigeria, Kenya and South Africa. Together they account for a significant share of innovation, venture capital, research output and advanced applications. Nigeria and South Africa were singled out as the continent's leading centres of AI research.
Between 2000 and 2024, Nigeria produced 4,428 AI-related publications with 64,925 citations, while South Africa produced 7,647 publications with 142,024 citations.
Those numbers place South Africa comfortably ahead on academic depth, reflecting decades of investment in universities such as Wits, Cape Town and Stellenbosch. Nigeria's figures, however, have been rising quickly, helped by a large and growing developer community and strong links between universities and the Lagos startup scene.
Adoption is wide but shallow
Perhaps the most striking survey result is that 44 percent of firms questioned in Nigeria and Kenya already use AI technologies. That compares with 61 percent among surveyed firms in the United States, a gap far smaller than many would expect. African businesses, it turns out, are quick to try new tools.
The catch lies in how those tools are used. The bank noted that sophistication remains lower in the two African countries, with many companies relying on basic applications such as chat assistants, translation or simple analytics rather than weaving AI into core operations like supply chain management, credit scoring or product design. Moving from experimentation to deep integration is where the real productivity gains lie, and it requires skills, data and reliable infrastructure.
The infrastructure gap
As Semafor Africa editor Yinka Adegoke highlighted, Africa is home to about 18 percent of humanity yet only around 0.6 percent of global data-centre capacity, and only about 5 percent of that capacity is equipped for advanced AI workloads. Global AI investment is expected to approach $2 trillion this year, according to Gartner, but very little of the cloud computing spending is landing on the continent.
Electricity is the underlying constraint. Data centres consume vast amounts of power and need it around the clock. In countries where grid outages are frequent, operators must invest in diesel generators, batteries or private solar installations, which raises costs and deters investment. Without stable, affordable energy, the trillion-dollar scenario stays theoretical.
Adopt and adapt, not build from scratch
The bank's central policy recommendation is pragmatic: most African economies should adopt and adapt existing AI systems rather than pouring scarce funds into building frontier models. Fine-tuning open-source models for local languages, regulations and business needs can deliver much of the value at a fraction of the cost. Initiatives in Ghana and Togo to build datasets in indigenous languages are examples of that approach in action.
Sector priorities are fairly clear. Agriculture, which employs a large share of the workforce, stands to gain from AI-powered weather forecasts, pest detection and market price information. Health systems could use diagnostic tools to stretch thin medical staff. Governments could improve tax collection, social payments and public procurement through better data analysis.
A warning about jobs
The optimism comes with a caution. On October 12, Vanguard reported that the World Bank and International Monetary Fund are both warning about potential job disruption in Nigeria and other African economies. Roles in customer service, data entry, basic accounting and translation are exposed to automation. With a very young and rapidly growing workforce, Africa can ill afford technology that eliminates entry-level positions without creating new ones.
Managing that transition will require serious investment in education. Coding bootcamps and university programmes help, but the bank's analysis implies that broader digital literacy, starting in secondary schools, is equally important so that workers across all sectors can use AI tools rather than be replaced by them.
What investors are watching
For investors, the report reinforces a familiar map: Lagos, Nairobi and Johannesburg remain the safest bets for AI startups, with Cairo, Kigali and Accra emerging as secondary hubs. Infrastructure plays such as data centres, renewable power for computing and undersea cable capacity are drawing growing interest from development finance institutions and Gulf sovereign funds.
The bottom line
The headline figure of $1 trillion will grab attention, but the more meaningful takeaway is conditional. Africa's AI dividend depends on electricity, data infrastructure, skills and smart regulation arriving together. Countries that coordinate those pieces over the next decade are likely to capture a disproportionate share of the gains, while those that do not risk watching the opportunity pass them by.