A record-breaking bond deal tied to Google-backed data centres has laid bare a widening gap between the pace of global artificial intelligence (AI) infrastructure investment and Africa’s capacity to compete for it, with power supply emerging as the continent’s most critical constraint.
The $5.7 billion offering, led by Morgan Stanley and linked to data centres backstopped by Alphabet’s Google, attracted $19 billion in investor orders when it priced on Thursday, April 16. The funds will finance two data centres in Sullivan County, Indiana, to be leased to cloud computing startup Fluidstack Ltd. The transaction is the largest dollar-denominated junk bond ever issued for a data centre project.
The sheer scale of private capital flowing toward AI infrastructure, and its concentration in power-secure locations, underlines what development economists and infrastructure analysts have flagged for years: electricity, not broadband or software talent, has become the primary filter determining where the AI economy takes root.
Africa Holds 0.6 Percent of Global Capacity
Africa is home to nearly one in five people on earth, yet the continent accounts for just 0.6 percent of global data centre capacity. The 2026 Data Centres in Africa Economic Report, published by the Africa Data Centres Association (ADCA), identifies power as the primary constraint on data centre growth, with grid instability in markets like Nigeria forcing operators to function as de facto independent power plants.
McKinsey projects that data centre capacity in Africa’s five largest markets will grow from around 400 megawatts today to between 1.5 and 2.2 gigawatts by 2030, with Kenya, Nigeria, Morocco, and Egypt identified as key emerging hubs alongside the continent’s leading market, South Africa.
But bridging that gap requires solving the power problem first. Nigeria, the continent’s most populous economy, has just 17 data centres collectively requiring around 137 megawatts of power capacity, and frequent outages remain a defining obstacle.
The Energy Economics of AI
The Google-linked bond reflects a broader structural reality. AI workloads are orders of magnitude more power-intensive than conventional computing, and the gap is widening as models grow in scale and usage. The International Energy Agency (IEA) projects that electricity demand from data centres, AI and crypto-assets could more than double by 2030.
Investors backing billion-dollar AI infrastructure projects are, in effect, making long-duration bets on jurisdictions that can guarantee a single input above all others: reliable, high-volume power. That calculus currently disadvantages most of sub-Saharan Africa.
Green Potential Versus Grid Reality
Africa’s renewable energy potential is widely recognised, but analysts are consistent in drawing a distinction between endowment and delivery. Ethiopia and Kenya already generate almost all their electricity from renewable sources, with Kenya drawing from geothermal, solar, wind, and hydropower. Microsoft and UAE-based tech firm G42 announced plans in May 2024 to develop a data centre campus in Kenya powered entirely by geothermal energy from the Olkaria field.
Those exceptions, however, underline the rule elsewhere. Grid reliability issues across much of the continent continue to push data centre operators toward diesel or gas backup generation, while water scarcity makes conventional cooling systems increasingly difficult to sustain. High upfront capital costs slow the adoption of advanced cooling technologies and energy storage.
The ADCA’s 2026 report notes that most African operators are currently adopting a modular, AI-provisioned design approach that allows graphics processing unit (GPU) capacity to scale as demand materialises, rather than committing to full hyperscale buildout in advance. The caution reflects both fiscal constraints and the infrastructure uncertainties that continue to deter the kind of long-term capital commitments the Google-linked bond represents.
The $5.7 billion transaction is therefore less a benchmark to replicate and more a measure of distance. Closing that gap will require the kind of sustained regulatory reform, grid investment, and public-private energy partnerships that have so far remained aspirational for most of the continent.
