BLUF: Heightened electricity demand from data centers could drive investments in Africa’s power systems, benefiting local economies and bolstering the revenues of utilities, while integrating the continent into global AI infrastructure. Yet, AI will not wait for the continent’s energy transition. A slow build-out could wind up leaving the region behind if core issues of grid reliability and regulatory certainty are not quickly resolved.
Relevance
- AI in Africa has been a major topic in recent weeks, with discussions at Davos 2025 and recent analyses highlighting its critical role in digital inclusion and economic growth, as well as the urgent need for grid infrastructure to support this transformation.
- Data center buildout in Africa has started, but overall investment is limited. Investment is pouring into new data centers across Asia and Latin America, yet Africa accounts for less than 1% of global data center capacity.
- New electricity demand in Africa is high, but new investment is constrained by lack of commercial offtakers. Data centers could provide stable income flows to unlock greater electricity infrastructure that would support broader economic activity.
Why consider investing in African data centers?
- The market is just taking off. Countries such as Egypt, Kenya, Morocco, Nigeria, and South Africa are beginning to attract data center investment. For example, the U.S. International Development Finance Corporation (DFC) invested $300 million and the International Finance Corporation (IFC) invested $250 million in Africa Data Centres, which is part of Lagos-based Liquid Intelligent Technologies Group and is the region’s first and largest network of data facilities. Additionally, California-based digital infrastructure firm Equinix recently announced plans to invest $390 million in South Africa.
- Huge untapped renewable energy potential. Technology companies seeking clean energy can find unexploited solar, wind, and geothermal sources across the continent. Uganda’s grid is 99% powered by renewables, while grids in Kenya and Ethiopia are 90% renewables-based. Microsoft’s partnership with KenGen will use geothermal energy for their new data center.
- Future market growth will be in Africa. By 2050, one-quarter of humanity will live in Africa, a massive market for digital services.
What’s in it for Africans?
- Data centers can support development. The high energy demands of data centers can incentivize improvements in local power grids and contribute to stable demand and revenue for utilities. This, in turn, can lead to more reliable electricity supply for surrounding communities, supporting broader industrial and development goals.
- Joining the AI race. Data centers are a prerequisite for Africa to join the global AI race and enable digital transformation to increase efficiency and unlock economic potential.
- Improving cloud services. Local data centers, such as Microsoft’s in Kenya and Google’s cloud region in Johannesburg, can reduce latency and improve the quality of digital services for users in Africa as demand grows. This lowers costs and enables new business.
- Jobs. Establishing local data centers can create direct and indirect jobs in construction, maintenance, and operations. Simultaneously, upskilling the tech labor pool will position African countries as competitive players in the global tech industry. Special economic zones for data centers like TRIFIC in Kenya are an early attempt to seize this opportunity.
What’s holding it back?
- Grid reliability. Nearly all African countries have outdated grid infrastructure and suffer from system losses (e.g., 16% system loss in Kenya and about 22% in Ghana)
- Grid bypass options. Potential economic spillovers to utilities and local business could be wiped out if data centers opt for private captive power solutions.
- Legal uncertainties. Fewer than half of African countries have approved or are in the process of developing AI strategies and regulations. Only 37 out of 54 countries have enacted data protection laws, while some governments insist on data localization which restricts cross-border data-sharing capabilities and, in some cases, gives them access to private data, undermining both investors and users. For example, JUMO, a British online banking platform, shut its operations in Rwanda due to restrictive data localization laws.
Conclusion
Africa’s energy sector and the potential for AI data center development are entangled in a chicken-and-egg dynamic: data centers need reliable, abundant power to operate, while improving power systems depend on stable demand from commercial customers like data centers. Africa has a choice: build the infrastructure to power AI or watch the future be built elsewhere.