IMF Paper Says Nigeria AI Gains Hinge on Reliable Power

In Short

An IMF paper says Nigeria’s near-term AI payoff is limited. The biggest blocker is reliable electricity, not the AI tools. Policy choices could widen the productivity gains range by 10x.

What Happened

A new IMF departmental paper, “Unlocking the Potential: Artificial Intelligence in Sub-Saharan Africa,” models what AI could add to economic output in the next decade.

The IMF estimates AI adoption could raise Sub-Saharan Africa’s productivity by 0.2% to 2.1% cumulatively over 10 years. Productivity is how much output an economy gets from its workers and machines. The wide range is driven mainly by policy, including infrastructure and skills, rather than by the technology itself.

Under current conditions, the IMF puts the region’s 10-year productivity boost at just 0.2%. That maps to a cumulative GDP increase of roughly 0.4%. It is lower than the roughly 1% gain the model projects for Europe and the Western Hemisphere under the same approach.

The paper also flags low diffusion, meaning how widely AI is being used. Using Microsoft data on AI-related activity on Windows devices, the IMF estimates AI diffusion in Sub-Saharan Africa at 9%. North America is at 30%, Europe and Central Asia at 22%, and the Middle East and North Africa at 21%.

Nigeria stands out because its job mix has higher “AI exposure,” meaning more roles include tasks that software can support or automate. The IMF lists Nigeria among the top countries in the region for potential AI-linked productivity gains.

Why It Matters

The IMF argues electricity is the binding constraint. About half the region’s population lacks reliable power. World Bank Enterprise Survey data cited in the paper says 78% of firms in Sub-Saharan Africa face routine outages, with average sales losses of 8.4% versus a 5.2% global average.

AI systems often need always-on computing to train and run models. Training is the compute-heavy process of teaching an AI model patterns from data, like repeated practice. Power cuts can interrupt training runs, slow response times, and increase costs.

In Nigeria, firms are adapting. The IMF notes that 86% of Nigerian firms own or share a generator, higher than comparable figures cited for Kenya and South Africa. That workaround keeps businesses running, but it also raises operating costs and can slow wider AI adoption.

For founders and investors building AI products in Nigeria, the message is practical. The biggest near-term lever is not just talent or data. It is dependable energy, plus the policy and infrastructure choices that make it cheaper to run compute at scale.