Everyone knows the story: artificial intelligence (AI) is the next big thing. Headlines focus on Nvidia (NASDAQ:NVDA) , Microsoft (NASDAQ:MSFT), and the handful of giants building the chips and software that power machine learning.
That's the easy story. The first-level trade. The one everyone sees.
But real money is made by asking what happens next. Second-level thinking involves examining the infrastructure that enables AI. The reality is that all this digital innovation depends on physical construction.
Those language models and neural networks live in warehouses filled with servers that require enormous amounts of power, water, cooling, and concrete. Every teraflop of AI compute rests on a foundation of steel and electricity.
The crowd buys the chips. The smart money buys the infrastructure. And AI's appetite for electricity is like nothing we have seen in modern history.
40,000+ Traders Use Benzinga Pro—Try It Free
See why serious traders refuse to trade without it. Real-time intelligence, AI analysis, and instant alerts. Test everything free for 14 days. Start Your 14-Day Free Trial
Industry analysts estimate that U.S. data center power demand will double over the next five years. By 2030, global data centers could consume as much electricity as the entire nation of Japan. That means more power plants, more substations, more transmission lines, and more engineers to build them.
Natural gas has become the backbone fuel of this new age. It offers reliability, scalability, and quick deployment. Companies like Cheniere Energy (NYSE:LNG), Kinder Morgan (NYSE:KMI), Williams Companies (NYSE:WMB), and Energy Transfer (NYSE:ET) are expanding capacity to meet rising demand near major data-center corridors such as Northern Virginia, Dallas-Fort Worth, and Central Ohio.
Nuclear power is also returning to the conversation. Southern Company (NYSE:SO) and Constellation Energy (NASDAQ:CEG) are among the first to benefit from new federal incentives and state policy tailwinds for carbon-free base-load generation. Renewables will remain an essential part of the mix, but even the most optimistic forecasts admit that wind and solar alone can't handle the load.
The message is simple. AI doesn't care where the electricity comes from. It just needs more of it, every hour of every day.
Once power exists, someone has to build the systems that deliver it. That's where the real opportunity lies.
Quanta Services (NYSE:PWR) is a dominant player in high-voltage transmission and grid modernization. The company has a record backlog of over $30 billion and continues to win contracts to expand power infrastructure for utilities, renewable developers, and data center operators.
Fluor Corporation (NYSE:FLR) is one of the world's largest engineering and construction firms, with a century-long record of tackling complex industrial projects. It builds everything from refineries and power plants to data centers and hydrogen facilities. Fluor's backlog recently approached $29 billion, with rising exposure to private-sector data center projects and power grid upgrades. The company doesn't pay much of a dividend right now, but management is focused on rebuilding profitability and returning more cash to shareholders in the coming years.
Jacobs Solutions (NYSE:J) provides the intellectual horsepower behind the physical buildout. Its engineers design cooling, power distribution, and structural components for data centers worldwide. Jacobs' dividend yield is about 0.8 percent, and the company's steady growth in infrastructure and technology consulting gives it a long runway of opportunity as the AI buildout expands.
Then there's Sterling Infrastructure (NASDAQ:STRL), a name most investors have never heard. Sterling handles the dirty but essential work that must be done before any data center rises — excavation, grading, underground utilities, and civil infrastructure that prepare the site. Its E-Infrastructure Solutions segment is now the company's growth engine, with a large, growing portfolio of data center campus projects across the Southeast and Midwest. Sterling recently announced a $100 million data-center project and expects roughly one-third of next year's revenue to come from this segment. Analysts see as much as 65% growth in Sterling's data-center business through 2026. Both its backlog and earnings growth are impressive.
Together, these companies form the engineering and construction backbone of the digital revolution. They are the ones literally building the new industrial landscape that will power AI.
Once built, those facilities have to be owned and operated. That's where the data-center real estate investment trusts come in.
Digital Realty Trust (NYSE:DLR) is one of the largest data-center REITs in the world. It owns and operates hundreds of properties across North America, Europe, and Asia, serving hyperscale clients such as Amazon, Google, and Meta. Demand for power-dense data-center space is outpacing supply, enabling Digital Realty to raise rents and maintain strong occupancy. The REIT currently yields around 2.7% and has a long history of annual dividend increases.
Equinix (NASDAQ:EQIX) is the other major global player in this space. It operates data center campuses that serve as digital interconnection hubs, allowing cloud providers, enterprises, and network carriers to exchange data securely and at high speeds. Equinix yields roughly 2.3% and continues to post double-digit growth in both revenues and cash flows as demand for high-performance data capacity soars.
These REITs may not have the highest yields, but they sit at the center of one of the most powerful secular growth stories of our time — the digital infrastructure that powers artificial intelligence.
The final piece of the puzzle lies in the utilities that generate and deliver electricity to this new digital grid.
NextEra Energy (NYSE:NEE) is one of the most innovative utilities in America, blending renewable energy leadership with traditional regulated utility operations. It yields around 3.5 percent and has a long history of dividend growth.
Duke Energy (NYSE:DUK) is another major player, serving the Southeast and parts of the Midwest. It is investing heavily in grid modernization, nuclear upgrades, and renewable capacity. Duke yields roughly 4.2 percent.
Dominion Energy (NYSE:D) serves the Mid-Atlantic, where data-center construction is booming. Dominion is expanding both base-load and renewable capacity to meet the growing demand in Virginia's "Data-Center Alley." The stock yields about 5.2 percent, making it one of the higher-income options in the group.
These utilities are the ultimate enablers of the AI economy. Without them, nothing else works.
What we're seeing is the convergence of two enormous capital cycles: the modernization of the American power grid and the global buildout of industrial-scale computing. Sterling builds the sites. Fluor and AECOM erect the facilities. Quanta connects them to the grid. NextEra, Duke, and Dominion provide the power. Digital Realty and Equinix lease the finished space. Nvidia and Microsoft fill it with hardware.
That's the entire chain from electrons to intelligence and every link offers opportunity.
Most investors are chasing the same trade. They buy the chips, the AI software names, and the flashy growth stories. But second-level thinking means going where the crowd is not. The infrastructure players may never make the cover of financial magazines, yet they are building the foundation that enables the entire AI revolution.
They pour concrete, wire substations, and connect grids. They collect dividends, lock in long-term contracts, and generate steady cash flow. For income investors and patient capital, this is the way to own the AI story without betting on the next technological fad.
AI may be the headline, but the real fortunes will be built by the companies turning electricity into intelligence — one substation, one trench, and one data-center shell at a time.
Editorial content from our expert contributors is intended to be information for the general public and not individualized investment advice. Editors/contributors are presenting their individual opinions and strategies, which are neither expressly nor impliedly approved or endorsed by Benzinga.
Photo: Shutterstock
