AM Law Alert Update : Sharing the Risk: Owners and Contractors Should Approach Data Center Projects Collaboratively
It’s not a secret; data center projects are a bright spot in the construction market. Going forward, the number of planned and estimated projects remains substantial. But while data centers look promising, should owners and contractors have concerns?
Well, yes. One recent study found that nearly 40% of data center projects expected to open in 2026 will be delayed by at least three months, and more than 60% of projects scheduled for 2027 have yet to break ground. Delays aren’t the only issue. According to some research, data center project cancellations more than quadrupled in 2025. There were six cancelled data center projects in 2024 and twenty-five in 2025.
For owners and contractors alike, the message is the same: the opportunity is real, but so is the risk that a project gets pushed, paused, or scrapped after labor, equipment, and capital have already been committed. The good news is that these risks are not mysterious. They tend to come from the same handful of sources, and parties who understand them can plan for them together. Below, we walk through the five recurring causes of time risk on data center projects, then turn to how owners and contractors can approach the negotiating table to handle them.
Why Data Center Projects Are Stalling
A construction project can be delayed for a variety of reasons. For data centers, however, it seems there are five common reasons.
Public opposition has gone mainstream. A March 2026 Gallup survey found that seven in ten Americans oppose building AI data centers in their communities, with 48% strongly opposed. To put that in perspective, that opposition exceeds what Gallup has ever measured for nuclear power plants in their area. Opponents most often cite resource consumption, particularly water and energy, along with pollution, traffic, and quality-of-life concerns. This sentiment is organized: some reports count at least 188 local opposition groups operating across 40 states. The political response has followed. Maine’s legislature passed a bill that would block large data center builds until late 2027, and similar moratorium proposals are under consideration in more than a dozen states.
Environmental litigation is becoming a serious roadblock. In May, a Minnesota court issued a temporary restraining order halting construction on Project Skyway, a 482-acre Google data center campus, after the Minnesota Center for Environmental Advocacy challenged the adequacy of the project’s environmental review under state law. The general contractor for the project estimates the delay could cost at least $5 million. Meanwhile, the NAACP is seeking a preliminary injunction against xAI’s Colossus data centers near Memphis, alleging that dozens of unpermitted gas turbines powering the facilities violate the federal Clean Air Act.
The labor pool may be shrinking as demand spikes. Roughly 41% of the current construction workforce is projected to retire by 2031, and nearly one in five electricians is already 55 or older. Some estimates indicate the U.S. power sector will need roughly 510,000 additional workers by 2030 to keep pace with demand. Construction executives on OpenAI projects have specifically pointed to a shortage of electricians and pipe fitters as a cause of delay.
Equipment lead times remain an issue. Long-lead electrical equipment is one of the (if not the) most cited on-the-ground bottlenecks. Medium-voltage switchgear is often running 40 to 60-plus weeks, with transformers and generators similarly delayed. As one data center developer put it, a shell can be complete and mechanically roughed in, but without the gear onsite, nothing gets energized.
A lack of infrastructure is commonplace. The grid in many regions simply cannot support data center facilities on the developer’s timeline. A site can look perfect on paper, but if the utility cannot commit to a firm delivery date, the project stalls. Hyperscalers have pledged to “build, bring, or buy all of the energy” their projects need, but utility interconnection timelines remain one of the most common reasons projects fail to launch.
A Shared Problem Calls for a Shared Approach
Here is the point that gets lost in a hard-fought negotiation: on a data center project, time risk is not the owner’s problem or the contractor’s problem. It’s the project’s problem. A moratorium, an injunction, a depleted labor pool, or a transformer stuck 50 weeks out will disrupt the schedule regardless of which party “owns” that risk on paper. And when the schedule slips, both sides risk losing money. That shared exposure is exactly why these deals reward collaboration over gamesmanship. The parties who do best are the ones who treat the contract not as a way to win the negotiation, but as a tool for managing a set of risks they both want to avoid.
So, how do owners and contractors do that?
Take reasonable positions on terms
A productive negotiation starts with each side recognizing what the other is, and isn’t, trying to accomplish. The owner’s objective should not be to punish a contractor for a delay the contractor did nothing to cause. Allocating the cost of a government moratorium to the party with no control over the legislature does not make the moratorium any less likely; it just sours the relationship and invites a dispute later. By the same token, the contractor’s objective should not be to convert every hiccup into a windfall. The aim is to be made whole for losses tied to risks outside the contractor’s control, not to profit from them.
When both parties come to the table with that shared understanding, the conversation changes. Instead of each side bracing for the other to overreach, the negotiation becomes a collaborative exercise in deciding, honestly, who is best positioned to bear or mitigate each risk. That posture tends to produce contracts that hold up better when something actually goes wrong, because the allocation reflects a genuine judgment about responsibility rather than a contest over leverage.
Be realistic (and maybe a little conservative)
Too often optimism drives pricing and bids, but results in delays and disputes. In the end, realistic expectations are what get projects built on time. For owners, that means understanding the marketplace pressures contractors are operating in. Labor and material costs are escalating, skilled crews are scarcer for data centers, and key equipment is on extended lead times. These are not bargaining tactics—they are well-known market conditions, and an owner who refuses to give them appropriate weight in pricing and scheduling is, in effect, buying a schedule that may not exist.
For contractors, realism means resisting the urge to win the work by promising more than the market can deliver. There appears to be no shortage of capital chasing this space, and owners and investors have shown a willingness to pay substantial sums for the right partner. But that willingness comes with the expectation that the contractor can actually perform. A bid built on labor or procurement assumptions the contractor cannot back up is not a competitive advantage; it is a delay claim waiting to happen. Building reasonable contingency into the schedule and the price protects both sides and keeps the project credible.
Use pre-construction to find the problems early
The most valuable time on a data center project is often the time before construction begins. Pre-construction is the parties’ best opportunity to surface the issues that cause delays later—gaps in the permitting path, uncertain utility timelines, equipment with long lead times that needs to be ordered now, labor availability in the specific market, and developing conditions (legislative, regulatory, or economic) that could reshape the schedule. A genuinely collaborative pre-construction phase, in which the owner and contractor map these risks together and coordinate around them, is the single most effective way to mitigate or eliminate time problems before they ever materialize. Risks identified early can be planned for; risks discovered mid-project become disputes.
Identify and specifically address the terms that govern delay, suspension, and termination
With the right mindset and good pre-construction work in place, the contract becomes the place to record how the parties have agreed to handle the things that can still go wrong. Both sides should pay close attention to the provisions governing delay, suspension, and termination, including the indirect ones that are easy to skim past but determine what actually gets paid.
A few questions are worth working through together. What compensation flows for a given delay? Does the owner cover demobilization and remobilization when work stops and restarts? Can the contractor recover idle or rental equipment costs during a suspension? The right answer to each depends on the cause of the delay and which party, if any, is responsible for it. Again, the guiding principle should be for the parties to be reasonable and realistic. Most of the time, that standard alone gets the parties where they need to be.
The harder work is filling the gaps a standard form leaves open. Off-the-shelf construction contracts generally do a competent job allocating delays caused by the owner or the contractor. What they do not necessarily and expressly cover are the five risks identified above. This is precisely why data center contracts should address them directly. If the legislature passes a moratorium that halts the project, is that a force majeure event, an owner-risk event, or something else entirely? Don’t leave it to be argued about after the fact; say so in the contract. The same goes for labor: if the available workforce is too thin, or if a competing project nearby siphons off the crews, how is that delay treated? Who absorbs the cost? Is there a way to work around the labor shortage, and what does it look like? Walking through each of the five risks discussed herein and addressing each directly can separate a contract that manages time risk from one that merely papers over it.
Takeaway
Data center construction remains a generational opportunity for owners and contractors alike. But the projects carry real time risk (community opposition, environmental litigation, labor scarcity, equipment delays, and power constraints) that can stall even a well-funded project. The parties who succeed in this market will be the ones who treat that risk as a shared problem: negotiating reasonably, planning honestly, and contracting deliberately around the issues most likely to disrupt the schedule.
For more information, please contact Adam Robertson.
Click here to receive updates in your inbox!
