While much of the commercial real estate sector tries to find its post-pandemic footing, data centers are sprinting into blue-sky territory. Record leasing activity, sustained hyperscaler demand, and accelerating AI investment have pushed the sector into an unprecedented phase of growth. This created what some described as a “couldn’t ask for more 2024,” followed by an equally stellar 2025.
Heading into 2026, data center fundamentals remain strong. Hyperscalers have signaled higher infrastructure budgets; operators report record backlogs; conversations around future deployments are already underway. From an investment perspective, data centers are maturing from a cyclical niche to a core element.
Record Demand, With More Still Ahead
Leasing and absorption across major US and European data center markets reached historic highs over the past year. Hyperscale tenants accounted for the majority of new capacity commitments, driven by expanding cloud services and the early stages of large-scale AI deployment.
While hyperscalers have not yet released formal 2026 capital expenditure guidance, commentary from recent earnings calls has been clear: Spending is expected to increase significantly and aggressively. That confidence is already filtering through the market. Data center landlords report that discussions around first-quarter 2026 deployments are already underway, a sign that demand is set to expand.
At the same time, new participants – often referred to as “neo-clouds” – are emerging as an additional demand source. These firms secure access to high-performance GPUs and lease large blocks of capacity to support AI workloads, which further tightens availability across key markets.
Power Access: The Primary Constraint
The main limit in today’s data center market is not tenant appetite but access to power. The markets forecast the greatest supply growth over the next several years, including Atlanta, Dallas–Fort Worth, Milan, Frankfurt, and Paris. These all share a common advantage: they can quickly bring large amounts of power online.
In contrast, some of the world’s most connected markets, such as London, Northern Virginia, Amsterdam, and Dublin, face challenges related to grid capacity and permitting timelines. Demand in these locations has not diminished, but the pace at which new capacity can be delivered has slowed.
This dynamic has accelerated growth in secondary and tertiary locations. In the U.S., West Texas, parts of the Midwest, and other rural areas are seeing gigawatt-scale pre-leasing activity – a level of demand difficult to imagine just a few years ago.
Latency Will Matter Again
That said, today’s geography of growth is not permanent. As AI applications evolve from training-intensive workloads to real-time inference – used directly by consumers and enterprises – latency will once again become critical. Data will need to live closer to its users.
This shift will not mean building data centers in dense urban cores, where land and power constraints remain prohibitive. Instead, it will favor secondary markets near major population centers that can balance connectivity, latency, and infrastructure availability. Over time, fiber networks will follow computing resources, reinforcing new hubs and gradually altering the relative importance of established markets.
Markets that aggressively restrict data center development may benefit in the near term from scarcity-driven pricing. But in the long run, they risk losing relevance as infrastructure investment migrates elsewhere. Encouragingly, some regions – Silicon Valley included – are recognizing the importance of planned, collaborative growth.
Why Investors See a Safe Haven
From an investment standpoint, data centers stand out for their return profile and strategic relevance. Stabilized net operating income yields exceeding 10% and development profit margins north of 50% reflect not only strong fundamentals, but also the sector’s higher complexity and execution risk.
A significant slowdown in AI monetization would have downstream implications for infrastructure demand. However, history suggests data centers will remain uniquely resilient. During recent periods of economic disruption, they outperformed many other real estate sectors as organizations leveraged digital infrastructure to control costs, automate processes, and maintain operations.
Looking ahead, the trajectory of AI adoption will remain a critical variable. Faster-than-expected monetization would likely drive even greater demand for capacity, while ongoing power constraints would further strengthen pricing for existing assets.
As the data center sector moves into its next phase, success will hinge less on identifying demand and more on understanding how technology trends, infrastructure capacity, and policy decisions intersect. Investors and operators who grasp those dynamics – and position themselves accordingly – will be best equipped to navigate what could be one of the most consequential periods in the sector’s history.
In the data center milieu, the combination of record demand, structural constraints, and long-term digital dependence is impossible to ignore. In an often unsettled commercial real estate landscape, this represents a kind of certitude no large language model could fault.
