Microsoft's ambitious $1 billion artificial intelligence data center in Kenya has stalled after government officials warned the project would require switching off "half the country" to meet its enormous power requirements, according to reports from infrastructure negotiations.
The impasse lays bare a profound tension in the global AI boom: as wealthy tech companies race to expand data center infrastructure worldwide, their massive energy demands increasingly collide with the fragile grids of developing nations—raising urgent questions about energy colonialism and climate justice in the AI era.
Kenya generates approximately 3,000 megawatts of installed electricity capacity, with significant portions already strained by existing industrial and residential demand. Microsoft's proposed data center would require power levels that officials described as untenable without catastrophic disruptions to national electricity access—a scenario that would disproportionately impact Kenya's poorest communities, who already face intermittent power.
The Kenya Energy and Petroleum Regulatory Authority emphasized that the country's grid infrastructure lacks the surplus capacity to support such intensive operations without major upgrades—investments that would take years and billions in capital that Kenya does not have. Microsoft has not publicly disclosed whether it would finance grid expansion, though such commitments are typically absent from data center negotiations.
In climate policy, as across environmental challenges, urgency must meet solutions—science demands action, but despair achieves nothing. Yet the Microsoft-Kenya standoff reveals how AI's climate costs are being externalized onto nations least equipped to bear them.
Data centers consumed an estimated 460 terawatt-hours globally in 2022, roughly 2% of global electricity—a figure projected to double by 2026 as generative AI accelerates. Training a single large language model can emit as much carbon as five cars over their lifetimes. These energy demands fall heaviest on the Global South, where tech companies seek cheap land and favorable tax incentives while exporting the environmental burden.
