Construction surges despite cost and power concerns
Construction is surging ahead despite cost and power concerns. Per Ermengarde Jabir, director of economic research at Moody's, municipal pushback and access to energy pose risks to future data center development, potentially affecting site selection and project timelines.
What happened
Data center construction spending continued its record-breaking pace in early 2026 despite mounting cost and power headwinds. Per ConstructConnect's June 2026 Data Center Report, year-to-date spending through April reached $49.5 billion, up sharply from $13.6 billion over the same period in 2025. Between October 2025 and March 2026 alone, $92.1 billion was invested across 140 new data center starts, with each of those six months ranking in the top 10 for monthly dollar spending on record.
Cost pressures
Costs are rising in step with demand. The trailing-12-month average data center project cost reached $475 million, more than double the $177.9 million recorded a year earlier. The median cost per square foot climbed to $445 in 2026, up 7.4% from full-year 2025, while the average hit $746 per square foot - a 45% jump year-over-year.
Power and site constraints
Power infrastructure starts rose 21.2% in Q1 2026 year-over-year, reflecting the electricity demands of new data centers. But energy access is becoming a gating factor. Construction Dive reporting, citing Ermengarde Jabir of Moody's Analytics, notes that municipal pushback and constrained power access are key risks to future development, particularly affecting site selection and permitting timelines. In a growing number of regions, local legislation is being introduced to slow large-scale data center growth.
Pipeline and geography
ConstructConnect is tracking 94 projects valued at $73.1 billion in the near-term preconstruction pipeline, heavily concentrated in the South - led by Texas, North Carolina, Virginia, and Arkansas. These projects are preconstruction-stage and not guaranteed to break ground, underscoring the gap between investment appetite and execution reality.
Significance for AI infrastructure
Data center construction is the primary physical layer of AI compute build-out. The surge in costs and the power access constraint are direct bottlenecks for cloud and hyperscaler AI capacity expansions. Community opposition and permitting delays introduce execution risk that affects timelines for bringing new GPU and inference capacity online.
Scoring Rationale
Solid data-driven construction industry report directly relevant to AI infrastructure build-out. The record pace of data center starts and the emerging power and municipal constraints are material signals for practitioners tracking AI compute capacity. Not a paradigm-shifting development, but practically important context.
Practice interview problems based on real data
1,500+ SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
