The risk to the AI economy: Engineering and construction labor shortage: PwC

Key takeaways

Billions in planned AI, clean energy, semiconductor, and infrastructure projects are competing for the same limited labor pool.
Workforce constraints are becoming a major risk to project timelines, costs, and investment returns across the AI economy.
Companies that secure construction capacity, invest in workforce development, and embrace productivity-enhancing technologies are better positioned to gain a competitive advantage.

Each week brings another headline. A hyperscaler commits $10 billion to a new data center campus. A utility announces a 2-gigawatt clean energy portfolio. Stock prices jump. Renderings circulate. Investors applaud. But no one asks a key question:

Who is going to build it? Institutional investors are allocating billions to infrastructure funds predicated on the assumption that a sufficient workforce exists to turn project commitments into physical reality. That assumption is wrong. The workforce shortage is getting worse with each passing quarter.

The investment community maintains quiet confidence that American ingenuity will sort this out. That confidence is not entirely misplaced. Markets do respond. Innovation does help. But these mechanisms operate on longer timescales than the ones investors are underwriting.

Investors committing capital to AI infrastructure, clean energy, grid modernization, and industrial reshoring would be well served to ask a simple question that is almost never asked in a pitch meeting: Who, specifically, is going to build this—and are they available?

The answer is, increasingly, no one.

This workforce crisis creates challenges affecting capital allocation decisions, project feasibility assessments, and operational execution capabilities. Understanding why this labor shortage is different is essential.

The scope of the workforce shortage

Baby Boomers and Gen Xers who’ve maintained America’s existing infrastructure are retiring in enormous numbers. They’re taking with them decades of institutional knowledge, field experience, and practical expertise that cannot be replaced. For thirty years, American culture and education policy systematically steered young people away from skilled trades. High schools eliminated shop classes. Vocational training declined. Trade careers lost ground to jobs that require four-year degrees.

The result? A generational gap that’s become a national economic vulnerability. The country faces a shortage of up to 1 million engineers, and about 40% of executives report difficulty hiring for critical roles.

The average electrical lineman in the United States is 52 years old. A full apprenticeship takes four years. The average hyperscaler wants power delivered in 18 months. That math does not work, and no amount of capital investment changes it.

Consider the numbers: More than 80,000 electricians will be needed each year over the decade, according to the US Bureau of Labor Statistics, but a small fraction of that number graduates each year from apprenticeship programs. About a half-million new construction workers will be needed to meet current demand.

In some regions, it takes six months just to get qualified crew on site.

Engineering tells the same story. The average age of a licensed professional engineer is climbing. Enrollment in civil, electrical, and mechanical engineering programs has been flat or declining at many institutions. The pipeline of engineers who know how to design a substation, route a transmission line, or engineer a foundation is thinning at precisely the moment demand is exploding.

The AI power surge

What makes the present moment different from previous construction cycles is not only the scale of AI investment, but the widening disconnect between the speed of software innovation and the physical realities of infrastructure delivery.

A single modern AI training cluster can consume as much electricity as a small city. The major technology companies have collectively announced hundreds of billions of dollars in data center investment over the next five years. US data center power demand could double or triple by 2030, requiring tens of gigawatts of new capacity.

The energy industry has spent years focused on the backlog of projects already waiting for grid connection approval. That queue is real. But it is increasingly a secondary bottleneck.

Even projects that secure interconnection agreements, financing, permits, and offtake contracts are discovering that the final obstacle is the most basic one: finding the people to build it.

At the same time, the economics of construction are becoming increasingly strained. A single 250,000 square-foot data center can now require approximately 1,500 skilled tradespeople to build. As hyperscalers, utilities, semiconductor manufacturers, and industrial projects compete for the same limited labor pool, wage inflation and contractor competition are driving substantial cost escalation and budget overruns.

And data centers are not the only source of demand:

The Inflation Reduction Act and Infrastructure Investment and Jobs Act are spurring clean energy, grid modernization, transportation, water, and broadband construction.
The CHIPS Act is catalyzing semiconductor fabrication plant construction.
The defense sector is expanding facility investment.
Industrial reshoring is generating new manufacturing construction nationwide.

The total backlog of announced but not-yet-built projects in the United States now represents something approaching a decade of construction activity at current workforce capacity. And all of this is competing for the same finite pool of electricians, pipefitters, ironworkers, heavy equipment operators, engineers, project managers, and construction supervisors.

Major technology firms have already begun adjusting expectations, with some already pushing back major data center completions by years, citing infrastructure constraints and delays in project execution.

For investors, the conversion rate—the percentage of announced projects that will be built on time and on budget—is almost certainly lower than what current valuations reflect.

What this means for you

Whether you're allocating capital, delivering projects, or building the teams that execute them, the workforce shortage is a concrete risk with immediate and actionable implications.

For investors

Identifying the blind spot

One of the most consequential aspects of this crisis may be that the investment community hasn’t fully priced it into their financial models.

The prevailing investment thesis for AI infrastructure is built on reasonable demand forecasts, policy tailwinds, and durable technology trends.

But there hasn’t been a realistic supply-side constraint analysis.

Workforce availability is an unfamiliar variable in financial modeling. Capital availability, regulatory risk, commodity prices, interest rates—analysts know how to hedge for them. But how do you account for a lack of qualified electricians to wire a data center in 18 months? It’s not a question that fits neatly into a discounted cash flow but could determine whether that data center generates revenue in 2027 or 2029. And a two-year difference will likely have a real impact on returns.

As a result, construction workforce availability is moving from a risk register line item to a critical path constraint that drives procurement strategy, project sequencing, and timeline commitments.

Strategies for infrastructure investors

Ask developers whether construction contracts, not just permits, are in place.
Determine if contractors who’ve committed to work can actually staff it.
Assess realistic delivery timelines given regional labor conditions.
Treat workforce availability as a due diligence essential, not an assumption.

For owners

Navigating the bid-delivery gap

When a developer releases an RFP on a $500 million data center or a $200 million transmission project, contractors will respond. Awarding work isn’t a challenge: seeing it come to fruition is.

The gulf between what contractors commit to and what they can deliver is widening. The warning signs are visible. At bid stage, contractors have thinner bench strength. Mobilization periods stretch. Subcontractors are lured away before finishing commitments. Supervision ratios are declining, given fewer experienced foremen. Quality drops, rework increases, and safety incidents rise. Claims and variations accumulate as schedules extend.

The most important question in contractor selection is no longer what is your price? It’s more fundamental: Will your people be available when we need them?

Procurement strategies for owners

Pre-commit to construction capacity before finalizing design.
Negotiate framework agreements that guarantee workforce allocation in exchange for volume.
Evaluate bids on demonstrated workforce availability, not just price.
Assess training pipeline investment and retention track records.
Treat construction labor as the scarce resource it is.

For engineering and construction companies

Weathering the workforce crisis with five practical solutions

No single solution closes the workforce gap in the timeframe the market demands, but several approaches can help. The contractors who adopt these aggressively may outperform. The owners who demand them are more likely to succeed.

Modular construction and offsite fabrication offer the single highest-impact near-term lever. Each hour that moves from a congested job site to a controlled factory gains productivity. This is particularly true for electrical rooms, mechanical skids, data center power modules, substation assemblies, and control buildings fabricated offsite. These require earlier design commitment and close coordination but can deliver measurable scheduling and labor advantages.

The average construction worker spends a significant part of each shift on non-production activities: waiting for materials, searching for information, travelling across site, reworking tasks due to unclear instructions. Tablets and other mobile devices can help improve front-line teams’ efficiency by providing quick access to current drawings, models, field workers’ task assignments, real-time progress tracking, digital daily logs, and automated material management.

Building Information Modeling (BIM) can offer significant value as a construction productivity tool. BIM allows teams to coordinate digitally before breaking ground. Use BIM to address spatial conflicts identified before construction begins.

Robotics and automated field operations are still nascent, but we’re seeing early adoption of specific applications. These include robotic total stations for autonomous surveying and layout, automated rebar tying and concrete finishing, and robotic welding for repetitive structural and piping connections. Drones increasingly perform inspections, monitor progress, and assess site logistics. AI-powered scheduling and resource optimization tools can improve decision-making, especially on projects with limited resources.

Many projects are staffed by workers from outside the local market. The projects that treat workers well attract and retain them. Those that don’t are likely to struggle with understaffing. Mastering workforce logistics can improve strategic capability. This includes:

Quality temporary housing and living facilities
Competitive per diem and travel allowances
Reasonable rotation schedules allowing workers to return home regularly
On-site amenities that make temporary assignments tolerable

Traditional four-year apprenticeships remain the gold standard in training new workers but the market currently can’t wait that long. The following can help:

Compressed training programs for workers with adjacent skills as well as military-to-construction transition programs leveraging existing discipline and technical aptitude
Industry-funded training partnerships with community colleges and technical schools
On-site training academies operated by major contractors
Virtual-reality training simulators allowing faster skill acquisition in safe environments
Modular credentialing that allows workers to become productive in specific tasks before completing full journeyman programs

A realistic approach to the future

These five solutions are not a one-size-fits-all approach. None of them, individually or collectively, fully closes the workforce gap. But they’re meaningful attempts to solve a persistent problem.

Broader structural responses will likely take longer. The industry requires dramatic scaling of apprenticeship capacity alongside accelerated pathways. It requires honest discussion of immigration policy and how it affects worker availability. Rising wages are attracting talent, but not at a sufficient scale. Investor due diligence should evolve to require workforce delivery plans alongside financial models.

In the meantime, developers, utilities, technology companies, investors, and policymakers should plan realistically for the workforce shortage. That means:

Longer timelines
Phased delivery
Earlier procurement of construction capacity
Honest communication about what is achievable and when

America is attempting something extraordinary: a simultaneous buildout of AI infrastructure, clean energy systems, grid modernization, semiconductor manufacturing, and transportation networks that collectively represent the largest physical construction program in generations. The capital is available. The technology is ready. The policy framework is broadly supportive. The investment community is eager.

And yet—while raising capital and announcing projects demands expertise—construction demands much more.

The great irony of the AI revolution may be that the most sophisticated technology in human history will ultimately depend on the availability of human hands willing and able to bend conduit and pull wire. The future of AI rests on the shoulders of electricians, engineers, and construction crews who have consistently risen to meet America’s grandest ambitions, from rural electrification to the interstate highway system. The question facing the AI economy is whether we can scale their expertise fast enough to match the unprecedented pace of digital transformation.

In the end, artificial intelligence may teach us the most human lesson of all: that no algorithm can replace the knowledge, dedication, and the craftsmanship of the people who construct the world we live in. The companies that act decisively today will likely weather this workforce crisis—and emerge as the builders of America’s next chapter.