By Staff Reporters
Published: Industry Insights & Infrastructure Report


Main Facts: The AI-Driven Infrastructure Gold Rush

The United States is on the precipice of an unprecedented structural transformation in its digital infrastructure. According to recent forecasts by Data Center Frontier, domestic data center capacity is expected to more than double over the next three years. Driven almost entirely by the relentless expansion of generative artificial intelligence (AI), machine learning models, and large-scale cloud computing, the industry is racing to deploy nearly 45 gigawatts (GW) of additional capacity.

However, this massive growth trajectory collides head-on with severe physical and logistical constraints. As outlined in JLL’s 2026 North America Data Center Midyear Report, supply is critically constrained. Prospective tenants are no longer able to easily secure sprawling, contiguous campuses; instead, they are forced to fight for small, highly fragmented blocks of space, with new leases routinely booked out for delivery as late as 2028.

These acute supply shortages have fundamentally altered market dynamics. Existing, stabilized data centers—facilities that already possess active, contractually secured power allocations—are experiencing a massive surge in valuation. Financial analysis released by Morningstar on August 5 highlights that these operational facilities are largely insulated from broader grid constraints and the increasingly common municipal interconnection moratoriums that are suffocating new construction pipelines.

Yet, older facilities face their own hurdles. Many legacy data centers lack the robust thermal management and structural load capacities required to support high-density liquid cooling systems, which are essential for powering modern AI Graphics Processing Units (GPUs) and hyper-scale compute clusters. Nevertheless, industry experts note that these sites can be systematically retrofitted to meet the moment.

To navigate these compounding pressures—ranging from power scarcity and grid bottlenecks to execution risks during the critical transition from construction to operations—industry players are forging new partnerships. A prime example is the recent market entry of L5 Mission Critical. By combining Walbridge’s century-long pedigree in constructing complex mega-facilities with L5’s advanced operational oversight, the joint enterprise aims to eliminate the friction points that traditionally plague new data center builds during their crucial first year of operation.


Chronology: How the Data Center Squeeze Developed

To understand the current infrastructure crunch, one must examine the rapid sequence of market shifts, strategic corporate acquisitions, and regulatory bottlenecks that have defined the commercial real estate and tech sectors over the past several years:

  • Pre-2022 (The Cloud Era): Data center development was heavily driven by enterprise cloud migration and hyperscale expansion. While power requirements were climbing, standard air-cooling methodologies and traditional electrical configurations were generally sufficient to meet demand.
  • Late 2022 – 2023 (The Generative AI Explosion): The public rollout of advanced generative AI models creates an overnight, insatiable demand for high-density computing. Power requirements per rack skyrocket, catching utility providers and developers flat-footed.
  • 2024 (Strategic Consolidations and Technical Services): Recognizing the impending operational crisis, major real estate and facilities firms begin consolidating technical expertise. Notably, JLL acquires SKAE Power Solutions in 2024, integrating its specialized electrical and mechanical capabilities into a newly formed technical services division under the leadership of veteran executive Matt Landek.
  • Late 2024 (Prefabrication Breakthroughs): In November, JLL strikes a landmark agreement with InfraPartners, a specialist in prefabricated AI data center solutions. The partnership aims to fast-track modular developments, attempting to bypass traditional, sluggish construction timelines.
  • August 2025 (The Midyear Reality Check): JLL’s 2026 North America Data Center Midyear Report confirms that capacity is severely constrained, pushing enterprise deployments deep into 2028 contracting windows. Simultaneously, Morningstar issues market commentary highlighting the immense shielding value of pre-existing, grid-connected facilities.
  • Present Day (The Build-to-Operate Pivot): Facing chronic labor shortages, operational handoff failures, and grid morasses, market leaders launch integrated build-to-operate models. Companies like L5 Mission Critical emerge to bridge the deadly gap between construction completion and live facility management, while simultaneously launching proprietary workforce training pipelines.

Supporting Data: The Metrics Fueling the Crisis

The numbers underlying the current data center landscape paint a picture of exponential demand clashing with unyielding physical limits:

  • 45 Gigawatts (GW): The estimated additional capacity planned across the United States to satisfy AI and cloud workloads over the next three years. To put this in perspective, 1 GW of electricity can power roughly 800,000 U.S. homes, meaning the data center sector is attempting to add the equivalent electrical load of 36 million households to the grid in a compressed timeframe.
  • Double in 3 Years: The projected multiplier for total U.S. data center capacity, according to Data Center Frontier.
  • 2028 Delivery Horizons: The current contracting timeline for major new deployments, illustrating that immediate, large-scale greenfield development is largely bottled up by grid interconnection queues.
  • 12-Month Vulnerability Window: The critical initial period during which new data centers historically experience the highest rates of downtime, operational friction, and stakeholder misalignment due to fragmented management handoffs.

Official Responses and Executive Insights

Industry leaders are speaking out about the systemic changes required to keep pace with the AI revolution. The shift from siloed construction and facilities management to holistic, integrated life-cycle partnerships is becoming a central theme among top executives.

John Rakolta III, president and chief administrative officer of Walbridge, emphasized the urgency of the current climate:

"As demand skyrockets, our clients not only need to get their projects up and out of the ground fast — they need critical operations and value-add services that reduce risk, ensure uptime and educate communities."

Rakolta also underscored the strategic leadership driving these new initiatives, pointing to Matt Landek’s background:

"Matt’s expertise will allow us to scale our data center construction practice to a full build-to-operate model that our clients need in this complicated time."

Matt Landek, drawing on his extensive experience managing complex portfolios—ranging from his tenure overseeing facilities management for the Cook County government in the Chicago metropolitan area to leading JLL’s technical services divisions—shed light on the operational blind spots plaguing the industry:

"L5 will differentiate itself by owning the handoff from construction and commissioning into live operations — the gap where most new data centers struggle through their first 12 months, and one no other provider has an incentive to close because each touches only a single link in the chain."

Landek elaborated on the structural philosophy of the new enterprise:

"We’re pairing that ownership with an operating model built for hyperscale from the start rather than retrofitted from legacy practice, a century of Walbridge experience running complex mega-facilities, and a workforce and community program that goes beyond what service providers typically offer."

Addressing the technical demands of modern hardware, Sean Farney, vice president of data center strategy at JLL, noted that while legacy facilities face hurdles, solutions exist:

"While many data centers may not be able to support liquid cooling that is required to support AI GPU storage compute, they can be upgraded or retrofitted to support this new demand."


Implications: Workforce Deficits, Grid Realities, and the Future of Facilities Management

The implications of this infrastructural bottleneck extend far beyond the balance sheets of tech giants and real estate investors. They touch upon regional power stability, municipal relations, and the future of skilled labor in America.

1. The Operations and Handoff Crisis

Historically, the data center industry has operated on a fractured supply chain model: real estate developers secure land, construction contractors build the shell and deploy electrical infrastructure, and third-party facilities management (FM) teams take over the keys upon commissioning. According to industry experts, this fragmented handoff is where the majority of facility failures occur during the first year of live operation. When high-density AI clusters go live, minor misalignments in thermal calibration or power distribution can lead to catastrophic downtime. Moving toward a "build-to-operate" model, as championed by ventures like L5 Mission Critical, represents a necessary evolution to mitigate these catastrophic risks.

2. The Workforce Skill Gap

Perhaps the most overlooked bottleneck in the data center boom is human capital. The specialized operational skill set required to manage high-density power systems, liquid cooling loops, and complex AI inference workloads far exceeds traditional facilities management training.

Matt Landek did not mince words regarding this deficiency:

"High-density power and AI inference workload management require a different operator skill set than most FM professionals were trained for. The handover from construction to the operating team is exposing new problem sets for operations teams not equipped for the demands and pressure of these operations. The market has not caught up on credentialing, training, or career pathways for these environments which is driving up demand for those experienced technicians."

To combat this, leading firms are beginning to build proprietary workforce training curricula. By creating accelerated, hands-on pathways for skilled tradespeople and investing heavily in advanced onboarding, proactive companies hope to solve the talent deficit before it throttles digital expansion.

3. Grid Constraints and Strategic Retrofitting

With greenfield developments bogged down by multi-year utility interconnection queues, the intrinsic value of existing, powered real estate will continue to soar. Companies that own stabilized sites will hold immense pricing power. However, property owners cannot simply rest on their laurels; they must aggressively audit their facilities for liquid-cooling retrofit viability. Those who successfully upgrade their legacy white space to handle dense GPU deployments will capture the windfall of the AI boom, while those anchored to outdated air-cooling paradigms risk obsolescence.

Conclusion

The next three years will define the winners and losers of the digital age. As the U.S. rushes to install 45GW of new capacity, success will not belong solely to those with the deepest pockets or the fastest construction crews. It will belong to the organizations that successfully bridge the gaps between construction and operation, modernize legacy assets for liquid cooling, and aggressively train the next generation of mission-critical engineers.

Leave a Reply

Your email address will not be published. Required fields are marked *