Most investors hear “AI” and think about semiconductor stocks, cloud platforms, or the latest language model making headlines around the world. Few stop to consider what keeps all of that running behind the scenes: raw electrical power and the infrastructure delivering it.
A new report from Charles Schwab argues that infrastructure companies sit at the center of one of the most compelling long-term investment themes. The opportunity stretches well beyond the tech stocks dominating your news feed, reaching into utilities, construction firms, and materials companies that underpin everything.
Your portfolio may already hold AI exposure through chip stocks, but you could be missing the physical foundation that powers the entire ecosystem. The real question is whether aging grids, surging electricity demand, and billions in government spending create a durable investment cycle or just another passing theme.
Schwab’s answer is clear, and the data supporting their position deserves your full attention before you make any portfolio decisions this year.
AI is creating an electricity crisis that most investors have not priced in yet
The demand for electrical power from AI data centers is growing at a pace that the existing grid was never built to support. Data center capacity needs could almost triple by 2030, with AI workloads driving the majority of that projected increase, according to McKinsey’s Data Center Demand Model.
U.S. data center power demand alone is expected to jump from 147 terawatt-hours in 2023 to over 606 terawatt-hours by the end of this decade.
“Global electricity demand from data centers is set to more than double over the next five years, consuming as much electricity by 2030 as the whole of Japan does today,” said Fatih Birol, IEA executive director.
Globally, $5.8 trillion is expected to be spent on grid upgrades between 2026 and 2035, with the U.S. accounting for roughly $1 trillion, according to JPMorgan.
Companies that build and maintain this infrastructure stand to benefit from years of sustained capital spending ahead of most other market sectors.
Data centers need power for training AI and for running it every single day
AI’s electricity consumption occurs in two distinct phases, and both contribute to the growing strain on power grids worldwide today. Training, where large language models learn to recognize patterns, requires enormous computational power concentrated over weeks or months at a time.
Inference, where trained models make real-time predictions on new data, runs continuously and consumes power around the clock. Electric vehicle adoption, the shift from gas furnaces to electric heat pumps, and changing weather patterns are all independently straining the grid.
The convergence of these forces is creating a structural deficit in electrical capacity that no single solution can address overnight or within one budget cycle. Demand for power in both AI phases is projected to keep growing through the end of the decade, according to Schwab’s research team.
The Charles Schwab research team says AI’s dual power demand is intensifying grid pressure, as training surges and constant inference drive global electricity use higher.
Governments are responding with historic infrastructure funding worldwide
The scale of spending needed to bring electrical supply in line with demand represents a generational investment cycle stretching over the next full decade. Germany approved a landmark 500 billion euro infrastructure fund in 2025, allocating resources over 12 years to modernize physical systems nationwide, according to Charles schwab.
That level of commitment signals how seriously developed nations are treating the infrastructure gap accumulated over decades of chronic underinvestment. In the United States, much of the electrical grid was built during the 1960s and 1970s, making large portions nearly 60 years old today.
Assets of that age need maintenance and replacement under normal conditions, but external threats from climate change to cyberattacks add urgent pressure.
“Grid resilience efforts are not just defensive infrastructure maintenance; they also increasingly underpin economic development, industrial competitiveness, and national security,” Sarah Kapnick, JPMorgan’s global head of climate advisory, wrote in the bank’s March 2025 report.
More AI Stocks:
- Morgan Stanley sets jaw-dropping Micron price target after event
- Bank of America updates Palantir stock forecast after private meeting
- Morgan Stanley drops eye-popping Broadcom price target
Global grid investment has accelerated in recent years due to rising power demand, according to BloombergNEF data cited by Schwab. The American Society of Civil Engineers rated U.S. energy infrastructure a D+ in its most recent assessment, highlighting the severity of the modernization challenge.
If you hold any broad market index fund, you already have indirect exposure to companies that will benefit from or be constrained by these realities. Global power demand is projected to grow at a 3.6% compound annual growth rate between 2026 and 2030, roughly 50% faster than the previous decade, according to JPMorgan.
All global data centers are projected to drive approximately 530 terawatt-hours of demand growth over the next five years, representing around 8% of total global electricity demand growth, according to the IEA.
For you as an investor, that translates into a sustained pipeline of capital projects that infrastructure companies will compete to build and service.
Infrastructure operators provide steady income and recession resistance for your portfolio
Not all infrastructure stocks behave the same way, and the distinction between operators and builders matters for portfolio construction decisions.
Operators include companies that own and run utilities, toll roads, airports, ports, and energy pipelines with long-lived assets already in place. These businesses offer several defensive characteristics that can protect your holdings during downturns and periods of heightened volatility, according to Schwab.
Key defensive characteristics of infrastructure operators:
- They provide essential services that remain in demand even during recessions and periods of slower economic growth across the broader economy.
- High barriers to entry and regulatory complexity create competitive moats that protect long-term profits from new competitors entering the space.
- Regulated frameworks and long-term contracts allow operators to pass rising inflation costs through to customers via rate-based pricing adjustments.
- Stable cash flows support consistent dividend payouts, which can be especially attractive during periods of broad market turbulence and uncertainty.
Builder-related stocks offer growth tied to multi-year capital spending cycles
If operators represent the defensive side of infrastructure, builders represent the growth opportunity tied to new construction and modernization programs. Builder-related companies include engineering firms, electrical component manufacturers, and materials suppliers producing the steel, copper, and cement that projects require.
These businesses are positioned to benefit from multi-year capital investment cycles driven by electrification, AI demand, and aging infrastructure replacement programs.
The four largest hyperscaler companies are forecast to spend nearly $650 billion in total capital expenditure. Approximately $450 billion is directed at AI infrastructure, a 71% year-over-year increase.
That spending flows to companies that produce generators, transformers, cables, fiber-optic components, and the heavy equipment needed for grid expansion projects. If you want exposure to the building phase of this cycle, these firms sit at the center of the capital deployment pipeline, feeding every new project.
How you can access infrastructure investments through ETFs and individual stocks
You have several options for gaining exposure to infrastructure companies through both individual stocks and diversified funds designed for this theme. Infrastructure ETFs generally spread holdings across the utilities, industrials, energy, and materials sectors to capture the full range of opportunities available.
Funds like the iShares Global Infrastructure ETF, Global X U.S. Infrastructure Development ETF, and SPDR S&P Global Infrastructure ETF are among the most widely held. Some funds also include technology companies that provide data center equipment and related components to the broader infrastructure buildout.
That exposure may add growth potential but also introduces more volatility than a pure utility or pipeline-focused fund would typically deliver. You can choose between U.S.-focused and globally diversified approaches, with international funds introducing foreign exchange considerations tied to dollar movements.
These risks could challenge the infrastructure investment thesis over time
Infrastructure investments carry specific risks you should evaluate before committing significant capital to any single strategy or fund in this space. Supply chain disruptions related to tariffs, geopolitical conflicts, and manufacturing dislocations could raise construction costs or create material shortages that delay critical projects.
Longer-term uncertainties include advances in quantum computing that could reduce energy per task and potentially slow the growth in demand. Regulatory and political pressures pose additional concerns for infrastructure investors evaluating a multi-year capital commitment to this sector.
Governments could impose price controls, windfall taxes, or consumer affordability directives that squeeze operators’ margins and reduce shareholders’ long-term income. Companies in the renewable energy industry may also face margin pressure from competition with Chinese government-subsidized exporters, adding another layer of risk.
The bottom line for your portfolio and your long-term planning decisions
Infrastructure presents a rare combination of defensive income and growth potential that few other asset classes can match in today’s market environment. By combining the steady cash flows of infrastructure operators with the capital appreciation potential of builder stocks, you can build a more balanced allocation.
These companies are not immune to economic slowdowns, but the secular forces driving demand for power extend well beyond any single business cycle. The AI boom everyone is chasing runs on electricity, and someone has to build the grid, generate the power, and maintain the systems delivering it daily.
If your portfolio is loaded with AI chip stocks but has zero exposure to the physical infrastructure supporting them, you may be missing the foundation. Schwab’s research suggests this is a multi-year theme with room to grow, and the window for positioning may still be open for patient investors.