Africa’s ability to participate in the artificial intelligence revolution hinges on a basic problem: the continent hosts less than one percent of global AI data centre capacity, even as its digital economy accelerates. That gap shapes what African businesses can build, what governments can process, and what ordinary citizens can access. A new venture between MTN Group and Dubai-based investor Tarek Al Ashram is positioning itself as a direct answer to that shortfall.
The partnership, called Africa Data Hub Holding, plans to build a 150-megawatt AI-ready data centre network across the region. The cost is substantial. Industry estimates place the price of 150 megawatts of AI data centre capacity somewhere between $3 billion and $6 billion, depending on technology choices, power infrastructure, location and deployment scale. The exact funding commitment has not been disclosed.
Additional reference context is available at https://businessday.ng/technology/article/mtn-dubai-tycoon-strike-deal-to-build-6bn-ai-data-centre-network-across-africa/?amp.
What makes this more than a commercial transaction is the dependency it aims to break. Without local AI computing capacity, African businesses, governments and developers must route their workloads through infrastructure located outside the continent. That arrangement increases latency, limits control over sensitive data, and restricts access to computing resources for African companies and public institutions alike. Local facilities could reverse those constraints directly.
Mastercard estimates Africa’s AI market could grow from approximately $4.5 billion in 2025 to $16.5 billion by 2030, a trajectory that will require significantly more computing power, cloud infrastructure, fibre networks and electricity. The continent already has more than 210 data centres, but most are conventional facilities designed to host websites and cloud workloads. AI data centres are a different category entirely: they require high-performance computing systems capable of training and running AI models, making them essential infrastructure for the next phase of digital development.
MTN has identified Nigeria and South Africa as the first deployment markets. Nigeria offers a large digital economy, a substantial consumer base and rapidly growing demand for cloud, fintech and enterprise technology services. South Africa has one of the continent’s most developed data centre and enterprise technology markets. MTN’s existing network in both countries provides a natural customer foundation for the new facilities.
Meanwhile, the operational challenges are considerable. AI data centres are extraordinarily power-intensive and cannot tolerate the prolonged outages that remain common across many African markets. MTN has already begun acquiring land and negotiating power agreements, groundwork that underscores just how much of this project depends on securing reliable electricity before a single server is switched on. Energy partnerships, captive power generation, renewable energy and battery storage are likely to become central to the project’s economics.
Al Ashram brings relevant experience to those challenges. As co-founder of Gulf Data Hub (backed by KKR & Co.), he oversaw large-scale data centre development in the Middle East. He has said Africa now presents an opportunity similar to the one that emerged in that region, driven by expanding digital economies and demand for resilient, scalable infrastructure.
For MTN, the venture marks a significant strategic shift. Under its Ambition 2030 strategy, the company is pursuing growth across connectivity, financial services and digital infrastructure. AI data centres fit directly into that framework, providing the computing power required for cloud services, digital payments, enterprise applications and government technology systems. MTN CEO Ralph Mupita has described a phased approach, beginning with 150 megawatts and using a partnership-led model rather than building alone.
The 150-megawatt target is likely only a starting point. MTN has indicated that future expansion will depend on demand, meaning the initial deployment could become the foundation for a much larger African AI infrastructure platform. Whether that expansion materialises at the pace and scale the continent needs is the question that will define this venture’s real public significance.