November 18, 2026
Balancing AI Data Center Demands with Energy Access in Africa

By Karikari Achireko, Director for Corporate Strategy & Global Partnerships, Africa Sustainable Energy Centre (ASEC)

Africa’s push to attract artificial intelligence investment is creating a new infrastructure challenge. While data centres can bring capital, cloud capacity, and digital services, they also require substantial investment in power generation, transmission, and grid connections.

According to the International Energy Agency (IEA), global data centre electricity consumption is expected to double by 2030, with the required infrastructure proving both expensive and long-lived. For African nations where power-sector capital faces competing demands and approximately 600 million people still lack electricity access this dynamic raises fundamental questions regarding cost allocation and risk distribution.

Serving large data centre campuses often requires new generation and transmission capacity, forcing utilities to invest before customers reach full electricity demand. This creates project risks, including construction delays, rising costs, and premature capital commitment.

Kenya’s Microsoft-G42 investment highlights these hurdles. Announced in 2024 as a $1 billion digital investment initiative centered on a geothermal-powered data centre and East Africa cloud region, the project later faced reported delays amid discussions over guaranteed annual payments and power requirements. Balancing a power system’s need for demand confidence with a developer’s need for guaranteed power remains a central policy hurdle.

Even when projects proceed, demand may materialize slower than forecasted, leaving infrastructure underused while costs linger. If utilities borrow to fund transmission upgrades without adequate customer revenue, financial burdens can fall on utility balance sheets, existing consumers via tariffs, or government guarantees.

To mitigate this, several African operators are taking direct responsibility for power supplies. In South Africa, Teraco is developing a 120 MW solar facility with electricity wheeled through Eskom, while Africa Data Centres signed a 20-year solar agreement with Distributed Power Africa. These models demonstrate that utilities and governments do not need to finance every incremental megawatt.

Effective contract design can allocate risks to parties best equipped to manage them. Measures such as connection charges, development milestones, minimum-demand commitments, and long-term power purchase agreements provide revenue certainty and protect utilities against stranded assets.

International models such as AEP Ohio’s strict tariff requirements or Ireland’s mandate requiring data centres to match import demand with generation/storage capacity—illustrate that large users can absorb an appropriate share of their operational costs and risks.

Protecting public balance sheets and consumers is only part of the equation. African economies must also evaluate what large digital loads deliver to the broader market. Comprehensive proposals like Microsoft-G42 extend beyond data storage into skills development and local AI capabilities, while international financing such as the IFC’s $100 million backing of Raxio’s carrier-neutral data centres proves that private capital can share the load.

Attracting AI investment should not require African power systems to absorb disproportionate risks. By establishing clear frameworks, connection commitments, and robust contract terms, African nations can transform AI infrastructure from a strain on electrical grids into a foundational anchor for broader economic development and digital capacity