Global spending on data centers is on pace to hit $31.6 trillion through 2050, an unprecedented capital commitment tied directly to the world’s expanding use of artificial intelligence. PricewaterhouseCoopers LLP laid out the figure in a report released Wednesday, describing a buildout that has no real historical parallel. The number itself is difficult to grasp: the cumulative figure is broadly comparable in scale to the annual gross domestic product of the United States, which is now above $30 trillion, although the two figures measure fundamentally different things. Should AI adoption move faster than PwC’s central scenario anticipates, the firm said total spending could climb as high as $50 trillion over the same period. Tech giants including Microsoft Corp. and Amazon.com Inc., alongside a growing roster of data center operators, are expanding computing capacity as demand for AI and cloud services rises across regions of the globe.

A New Category of Capital Expenditure

PwC frames the current AI infrastructure cycle as something categorically different from the great capital projects that came before it. Railroads, electrification and the early internet each demanded enormous sums and, in their own time, defined the economic era around them. “Railways. Electrification. The internet. Each required enormous amounts of capital and defined an era,” the researchers said in the report. “The AI infrastructure cycle underway dwarfs all three. This one resets every four to six years — and shows no signs of ending.” That four-to-six-year reset cycle is the detail separating this buildout from its predecessors, pointing to a spending pattern that renews itself rather than tapering off once initial construction wraps. Unlike a rail network or the physical infrastructure of a fiber route, which can have long useful lives once built, a data center’s technology layer ages out of usefulness within a handful of years and requires regular replacement.

Where the Money Is Going

The United States is set to capture the largest share of the projected spending, at $15.1 trillion, or nearly half of the global total, PwC found. Asia-Pacific follows at $8.2 trillion, positioning the region as the second-largest theater for data center investment through 2050. Europe trails at $5.6 trillion, while the Middle East and Africa round out the picture at $1.1 trillion and $255 billion, respectively. On an annual basis, the growth trajectory tells a similarly steep story: global data center spending will rise from roughly $800 billion this year to $1.1 trillion by 2030 and $1.8 trillion by 2050. China and India, buoyed by large populations and rapidly digitizing economies, will drive the largest share of that incremental demand. Both countries carry substantial headroom for AI to embed itself deeper into everyday business and consumer activity, a factor PwC treats as central to its long-range projections.

PwC commissioned Oxford Economics Ltd. to build the underlying model powering these figures, and the resulting report spans 46 countries and territories across five regions. Those regions together account for the bulk of global economic activity and digital infrastructure investment, giving the forecast a genuinely worldwide scope rather than a narrow, US-centric view. The breadth of the dataset matters because data center economics vary sharply by geography, shaped by everything from electricity prices to land costs to regulatory posture. Researchers built the model to capture those regional differences rather than applying a single blanket assumption across every market. The result is a picture in which the United States dominates in absolute dollars, yet growth rates in Asia and parts of the Middle East may ultimately outpace the mature American market.

The Replacement Economy Behind the Numbers

What makes this spending cycle unusual is not the initial construction of a facility but what happens after the building is finished. Graphics processing units, servers, storage systems and networking equipment all require routine replacement, and PwC’s researchers expect recurring hardware upgrades to account for the majority of total investment. That stands in sharp contrast to earlier infrastructure cycles such as railways, electrification and the internet, which front-loaded capital expenditures and tailed off as the networks matured. Nvidia Corp., the leading supplier of AI accelerators, sits at the center of this replacement economy because GPUs are among the key pieces of ICT equipment that must be refreshed as computing requirements evolve. As computing demands intensify, the recurring replacement of chips and other ICT equipment will continue to pull capital into successive upgrade cycles that might otherwise have been spread across a longer horizon.

Power, Sovereignty and the Global Chip Chess Match

Strong global demand for AI infrastructure does not guarantee that every region will benefit equally, and PwC identifies power availability as the foremost factor determining where investment ultimately lands. Affordable, reliable and increasingly low-carbon electricity at scale presents the hardest requirement for many markets to meet, according to the firm. Data sovereignty rules and the flow of semiconductors across borders will further shape which countries pull ahead in the competition for capital. PwC’s central forecast assumes a relatively open trading system in which chips move freely worldwide; however, disruptions to semiconductor supply chains could cut global investment by nearly 20%. Meanwhile, PwC expects a growing push toward data sovereignty to redistribute investment across borders while modestly reducing the overall global pool of capital.

“The $31.6 trillion question isn’t whether the capital exists. It does,” the researchers said. “Nor is the question whether the demand is real. It is. The question is which regions, operators, and institutions are positioned to capture it and which aren’t.” That framing cuts to the heart of the entire report, since PwC expects sufficient capital to meet the buildout while identifying demand growth, power availability, chip access and other infrastructure constraints as key variables shaping the cycle. By PwC’s own account, the open variables include execution: power and grid capacity, permitting and policy certainty, access to advanced chips, connectivity, security and community consent. Markets that make data center development faster, more predictable and more investable are more likely to capture a larger share of the buildout.

The Backlash Data Centers Can’t Ignore

Data Center Watch found that local opposition blocked or delayed at least 75 projects worth about $130 billion combined in just the first three months of this year. Communities pushing back cite a familiar set of concerns: environmental impact, strain on local resources, and broader anxiety over how AI could reshape employment and daily life. Local zoning fights and permitting battles have already forced some operators to redesign projects or relocate them to friendlier jurisdictions altogether. In response, developers and local authorities are increasingly turning to community outreach, local hiring commitments and greater transparency around water and power usage in an effort to address opposition before it escalates. Whether these efforts prove sufficient could influence how much of the projected $31.6 trillion in spending ultimately reaches markets where developers can secure the local support and approvals needed to proceed.

What Happens Next

The scale of PwC’s forecast leaves little doubt that AI infrastructure has become one of the defining capital stories of this century, comparable in ambition to the great infrastructure booms that preceded it. The report makes clear that capital should be available in sufficient quantities, while securing the electricity, infrastructure and other conditions needed to deploy that capital remains a major challenge. The harder task lies in securing the electricity, chips and social license needed to actually deploy that capital at the pace the forecast assumes. Countries that can make data center development faster, more predictable and more investable are better positioned to attract a larger share of the buildout, while markets constrained by power, planning or policy hurdles risk losing investment to more competitive jurisdictions.