A policy signal emerging from the United States this week offers a revealing preview of a debate Africa has not yet fully confronted: who pays to power the digital economy.
According to reporting by Bloomberg, US President Donald Trump, working with governors across the US Northeast and Midwest, is expected to direct PJM Interconnection to hold an emergency power auction, to compel large data center operators to underwrite the construction of new electricity generation capacity needed to support their rapid expansion.
If implemented, the mechanism would allow hyperscalers and technology firms to bid for 15-year power contracts tied directly to new generation assets, potentially unlocking more than $15 billion in investment. It is a tacit admission that traditional grid planning models are no longer sufficient for AI-scale computing, and that the public sector is increasingly unwilling to socialize the costs of private digital growth.
The policy move follows President Trump’s public insistence that large technology companies, beginning with Microsoft, must “pay their own way” when it comes to the power infrastructure required for data centers. Already, power demand from data centers across the PJM market is accelerating faster than generation can be financed, built, and connected. PJM forecasts peak load growth of 3.6 percent annually through 2036, largely driven by data center development, with mounting concern that residential and industrial ratepayers could ultimately bear the cost through higher electricity bills.
For Africa’s regulators, investors, and policymakers, this episode should land as more than foreign news. It is an early warning.
Across Africa, data center pipelines are expanding rapidly – from Lagos and Johannesburg to Nairobi, Cairo, Casablanca, and Accra. Governments are courting hyperscalers, cloud providers, and colocation developers as anchors of digital transformation, fintech growth, AI readiness, and data sovereignty. Yet in most markets, power planning remains structurally decoupled from digital infrastructure planning. Data centers are treated as just another large customer, rather than as quasi-utilities with long-term, predictable, and concentrated demand profiles.
The US debate underscores a strategic inflection point Africa will soon face. As compute density rises and AI workloads proliferate, the question will no longer be whether power is available, but who finances the incremental capacity, how risks are allocated, and whether grid users are shielded from cross-subsidization. Markets that fail to resolve this early risk deterring investment or importing an unsustainable political backlash once electricity tariffs begin to reflect digital demand.
There is, however, a countervailing opportunity. Africa is not yet locked into legacy grid structures. Power-anchored data center campuses, dedicated generation, long-term power purchase agreements, gas-to-power solutions, and renewable-hybrid models are already being, and will continue to be structured as part of data center investment theses. In doing so, Africa can leapfrog toward a model where digital infrastructure growth directly catalyzes new power assets, rather than quietly competing with households and industry for scarce supply.
Beyond asserting dominance over hyperscalers, the emerging US approach is about realigning incentives so that the builders of digital capacity also help build the physical systems that sustain it. African markets would do well to internalize that lesson now, before power becomes the binding constraint on its digital ambition.
