IMF: AI could raise sub-Saharan Africa GDP by up to 4% over a decade if infrastructure barriers fixed

Artificial intelligence could increase sub-Saharan Africa’s gross domestic product by about 4 percent over the next decade if countries address key barriers such as electricity supply, internet access and digital skills, according to a new International Monetary Fund departmental paper. Without those reforms, the economic payoff would be minimal, on the order of 0.2 percent of GDP, described by researchers as little more than a rounding error.

The paper, titled Unlocking the Potential: AI in Sub-Saharan Africa and prepared by an IMF African Department team led by Martin Schindler and Andrew Tiffin, was released in July 2026. It applies AI exposure and productivity methods to the region and models different adoption scenarios. Under current conditions of low adoption rates, a large share of employment in low-exposure sectors such as agriculture, and limited complementary inputs, productivity gains are estimated at just 0.2 percent over 10 years. With faster diffusion, broader sectoral penetration and improved enabling conditions, productivity could rise by about 2.1 percent and cumulative GDP by nearly 4 percent—equivalent to roughly half a percentage point of additional annual growth, the analysis shows (IMF paper; Reuters).

“Policy changes will be key to whether further growth can be unlocked from AI,” Schindler, deputy division chief and mission chief in the Fund’s African Department, told Reuters. Without decisive action, many countries “may see productivity and growth gains of just 0.2% over the next decade... Frankly, that’s a rounding error,” he said (Reuters).

Sub-Saharan Africa ranks lowest on the IMF’s AI Preparedness Index, reflecting shortfalls in digital infrastructure, technical skills and regulatory capacity. Around half the region’s population lacks reliable electricity, and only 38 percent of Africans used the internet in 2024 compared with a global average of 68 percent. Africa hosts roughly 160 data centres, about 5.5 percent of the global total, with nearly half concentrated in South Africa, Nigeria and Kenya (Reuters; Down To Earth).

The central concern for the region “is not the risk of technological disruption, but whether countries will be able to adopt, adapt, and scale AI quickly enough to capture its benefits and avoid falling further behind,” the paper states. Co-author Andrew Tiffin noted that electricity is foundational: “It’s hard to have anything without electricity.” Data centres could themselves become bankable projects that accelerate electrification (Reuters).

Rather than primarily displacing office workers, AI’s main promise in Africa lies in raising productivity across the economy—helping informal firms manage inventory, enabling farmers to increase yields, supporting education and healthcare delivery, and improving public-sector functions such as tax compliance. Early examples include chatbot tutoring pilots in Nigeria that delivered learning gains and digital agricultural advisory services in several countries that improved crop yields when combined with better inputs (Punch; IMF paper).

An accompanying IMF Country Focus blog emphasises two priorities: building foundations through investment in power, broadband, data infrastructure and digital literacy; and establishing clear rules on data, competition, cybersecurity and public-sector AI use to build trust. Regional cooperation is also highlighted, as many economies are too small to develop competitive AI ecosystems alone. Private investments already underway include a Microsoft-G42 geothermal data-centre project in Kenya and a Cassava Technologies-NVIDIA GPU deployment across several countries (Punch; Reuters).

The paper concludes that AI is central to the region’s growth strategy. “Africa does not need to win the race to build cutting-edge AI models, but it must find ways to use AI widely, cheaply, and safely,” it notes. The outcome over the next decade will depend on choices made by governments, schools, farms and firms across the continent (Punch).