Everyone is talking about Africa’s AI opportunity, but very few are talking about what will actually power it. AI may be built on algorithms, data and clever machines, but none of it gets very far without a reliable supply of electricity.

Africa has the population, the digital appetite and the potential to attract hyperscale investment, but if it cannot produce and deliver enough affordable power, the continent’s AI ambitions risk becoming another impressive strategy document. AI doesn’t run on hype. It runs on megawatts.

The scale of the opportunity is already becoming clear. McKinsey estimates that demand for data centre capacity in Africa could rise from about 0.4 GW today to between 1.5 GW and 2.2 GW by 2030, requiring between $10 billion and $20 billion in new investment in data centre shells alone.

That is a remarkable opportunity for a continent that currently accounts for only a fraction of global data centre capacity. The problem is that every additional megawatt of computing capacity creates a corresponding question about where the electricity will come from.

The International Energy Agency‘s latest 2026 analysis makes the global direction unmistakable. Electricity consumption from data centres increased by 17 percent in 2025, while consumption from AI focused data centres grew even faster.

Globally, data centre electricity consumption is expected to roughly double from 485 TWh in 2025 to around 950 TWh by 2030, with AI focused facilities expected to triple their electricity consumption over the same period.

The technology industry may be moving at extraordinary speed, but electricity infrastructure does not. A data centre can be constructed in a few years. Generation, transmission and grid infrastructure require much longer planning cycles.

That creates a particularly awkward challenge for Africa. Around half of the population of Sub Saharan Africa still does not have reliable access to electricity, according to the International Monetary Fund (IMF). At the same time, the continent is being asked to support an entirely new class of electricity intensive infrastructure.

This is not simply a choice between powering homes or powering data centres, it’s a question of whether investment in generation, transmission and distribution can expand quickly enough to do both.

The IMF estimates that AI could raise economic output in Sub Saharan Africa by as much as 4 percent over the next decade if the region improves electricity, connectivity and digital skills. Without those improvements, the economic gain could be dramatically smaller.

This is where the conversation needs to change. Data centres should not be viewed solely as enormous consumers of electricity. They can also become anchor customers for new generation. Their demand is large, predictable and long term, which can help make new renewable projects, transmission investment and private power arrangements more commercially viable.

The opportunity is to build energy and digital infrastructure together rather than waiting for one to catch up with the other. Power developers need customers. Data centres need power. Africa needs both. That is an unusually useful alignment of interests.

South Africa illustrates what could happen when those pieces begin coming together. It already has the continent’s most mature data centre ecosystem, extensive fibre connectivity and established cloud infrastructure.

The country is also changing the way electricity can be generated and procured. Renewable energy projects, wheeling arrangements, private generation and power purchase agreements are creating alternatives to depending entirely on traditional utility supply.

The question is no longer simply whether South Africa has enough electricity today, it’s whether the country can create enough new, reliable and competitively priced electricity to support the next generation of AI infrastructure.

Other African markets have opportunities of their own. Countries with abundant solar, wind, hydro or geothermal resources could attract data centre investment if they combine those resources with reliable grids, international connectivity, suitable land and predictable regulation.

Kenya, for example, has already attracted significant attention because of its geothermal resources and growing digital ecosystem. South Africa has its solar and wind potential. Namibia has enormous renewable energy ambitions. Egypt has strategic connectivity and growing data centre investment.

The winners will not necessarily be the countries with the biggest existing technology markets. They may be the countries capable of delivering megawatts quickly, reliably and at a price that makes hyperscale investment commercially attractive.

That makes electricity policy an AI policy. Governments need to stop separating digital infrastructure strategies from energy strategies and start planning them as one economic system. Regulators need to make grid connections, land approvals, power procurement and infrastructure licensing predictable. Investors need to look beyond the data centre itself and consider the generation, transmission, storage and connectivity infrastructure around it.

Operators, meanwhile, need to think differently about where they build and how they secure long term power. Africa’s AI race will not be won by whoever has the most impressive technology announcement. It will be won by whoever can keep the servers running.

The continent has a rare opportunity to turn its energy transition into a digital infrastructure advantage, but the window will not stay open indefinitely. Africa does not need to wait for the AI economy to arrive. It needs to build the electricity infrastructure that will allow it to stay.