How AI’s Hunger for Power Could Help Electrify Africa

Across sub-Saharan Africa, 560 million people don’t have access to electricity. In cities, towns and villages across Africa, communities and businesses have to operate without reliable power, or rely on dirty and dangerous kerosene lamps and expensive diesel generators.

Meanwhile utilities and technology companies worldwide are racing to expand power infrastructure to meet growing demand for artificial intelligence (AI). In the United States and elsewhere, that expansion has become a political hot potato due to concerns about electricity prices, stress on the electrical grid, and other community impacts like noise and water.

But what if AI’s insatiable appetite for electricity could help bring inexpensive, reliable, and sustainable electricity to African communities, enabling businesses to grow and lifting people out of poverty?

That is the question that the Emergent Grid Project is trying to answer.

When I served as coordinator of Power Africa, the U.S. Government-led initiative to double access to electricity in sub-Saharan Africa, from 2013-2020, I saw firsthand how many African energy systems lacked sufficient power generation. One reason was that customers could not afford the commercial rates needed to keep the energy system online and financially sustainable. Now, as CEO of the Mission 300 Accelerator, I believe the growing demand for AI may offer a surprising solution.

Not every AI request can be processed across the world, but many computing tasks can run wherever reliable power and connectivity exist. So, someone in Atlanta might use AI to plan a vacation or generate an image of their new kitchen design, with “compute” – the processing power needed to run AI and other computing tasks – purchased from a mini-grid in Sierra Leone. That transaction creates a new source of income for the mini-grid, helping finance reliable and affordable electricity for the surrounding community.

Instead of driving up the cost of electricity, the global demand for digital compute could make electricity available, reliable, and affordable for people anywhere in the world. That’s the Emergent Grid: compute powered from anywhere in the world.

The Emergent Grid

The Emergent Grid marks a shift from centralized grids and massive data centers toward a digitally connected network of smaller, modular computing facilities located close to available energy.

Many rural African communities, such as Kasiri in Sierra Leone, rely on heavily subsidized mini-grids: self-contained systems, some as small as a shipping container, often built from solar panels and batteries. Yet in Kasiri, the solar mini-grid leaves electricity generation unused because output peaks during the day while household demand peaks at night, while many rural households simply can’t afford the power they do use.

The Emergent Grid could change that math. Companies running AI workloads could buy the excess daytime electricity when community demand is low – gaining compute capacity in the process – while providing mini-grids with a predictable revenue stream. Over time, that could reduce mini-grids’ reliance on subsidies and help finance power projects for households and businesses alike.

Testing the Concept

Working alongside philanthropies, developers and technical experts, the Mission 300 Accelerator, housed within The Rockefeller Foundation’s charitable offshoot RF Catalytic Capital (RFCC), is launching the Emergent Grid Project on the margins of the 2026 UN General Assembly to test whether the model can improve the economics of African power projects.

Four principles guide this project. First, communities come first: households, businesses, schools and health clinics always have priority, with compute strengthening local access rather than competing with it. Second, compute is a tool, not the objective; the goal is reliable, affordable electricity. Third, digital demand gives mini-grids a reliable anchor customer and steadier revenues while local demand grows. Finally, we will test before we scale, building evidence about whether, where and how this model delivers benefits and can be responsibly replicated.

The opportunity could extend beyond electricity. Since many computing workloads require reliable connectivity as well as power, the project will explore opportunities to leverage that connectivity to expand digital access and economic opportunities for surrounding communities.

We will work with experienced mini-grid developers to add digital loads to sites, including in Sierra Leone, Kenya, and the Democratic Republic of Congo, alongside independent monitoring and evaluation to assess the model’s performance and community impacts.

If the project succeeds, having reliable electricity and digital connectivity would allow businesses to start, grow, and hire. Farmers could process and refrigerate crops and use AI to help diagnose crop diseases. Merchants could power lights to operate at night and use AI to conduct market research and efficiently manage inventory. The resulting activity would raise incomes and create jobs in communities long excluded from modern energy and digital economies.

This could be a game changer for rural electrification, a key priority of Mission 300, a partnership led by the World Bank Group and the African Development Bank Group, which aims to connect 300 million Africans to electricity by 2030. Mission 300 has already brought electricity access to more than 50 million people, with roughly half of the remaining connections expected to come through grid expansion and the rest through mini-grids and solar home systems.

Grants and subsidies alone are no longer enough to bring electricity to everyone. By connecting energy access to the rapidly expanding global market for AI, Africa could turn AI’s enormous appetite for electricity into an engine for electrification, business growth, higher incomes, jobs and returns for investors.

AI’s insatiable appetite is viewed as a problem; in the right places, it could be the solution.