Most Americans only understand the purpose of a data center as a means to power AI. That’s an understandable perception, since chatbots, image generators, and AI-powered assistants have dominated headlines and driven an historic wave of data center construction. But that headline obscures a much bigger story. Long before the most current AI boom, data centers were already the invisible infrastructure behind everything Americans do online.
A data center is, at its simplest, a building full of servers that store, process, and deliver information. When someone sends an email or text message, saves a family photo to the cloud, checks a bank balance, streams a football game, or looks at their health records, that request travels to a data center for processing and comes back to the device in about a second. Data centers are the backbone of the entire digital economy.
The Everyday Services Running on Data Centers Right Now
Consider how much of a typical day most Americans spend online. One national survey found the average American now spends more than ten hours a day connected to the internet, using it to browse, stream entertainment, pay bills, and peruse social media. More than a third log on to handle work assignments and one in five use it for schoolwork.
That schoolwork increasingly runs through the cloud: Google Classroom alone supports more than 150 million students, educators, and school leaders worldwide, turning homework, grading, and parent-teacher communication into an always-on digital exchange. Add in the record 132 trillion megabytes of wireless data Americans used in 2024 alone, and a clear picture emerges. Whether it’s a parent paying a bill, a kid submitting an assignment, or a quick text to confirm dinner plans, virtually none of it happens without a data center working behind the scenes.
Streaming video alone accounts for the majority of all online traffic, representing over half of all data moving across the internet. Banking and finance depend just as heavily on the expansive online infrastructure needed to power mobile banking apps, real-time trading platforms, and fraud detection systems that protect customers around the clock. Mobile banking is now the primary way U.S. households access their accounts, a nearly ninefold jump over the past decade.
Healthcare has undergone a similar shift: hospitals and clinics now store patient records digitally, making it possible for a doctor to see a patient’s history instantly rather than waiting on a paper file to be located and faxed across town. E-commerce depends on data centers to process transactions, manage inventory, and keep checkout pages loading during even the busiest shopping days of the year.
The Common Misconception: “Data Centers Are All for AI”
The confusion is easy to trace. Recent data center construction has surged specifically because AI workloads demand more computing power per server rack than traditional data centers, and global data center capacity will nearly double by 2030 to meet projected demand. Because AI-driven construction is what’s making headlines and drawing new investment into states, it’s easy to assume AI is the reason data centers exist, but that gets the story backward.
The facilities being built today extend the infrastructure that is already running the internet. Virginia alone hosts hundreds of data centers that predate the current AI investment wave, built originally to support cloud storage, telecommunications, and enterprise computing.
The Stakes of Allowing the Misconceptions to Continue
Why does this distinction matter? Because policy debates about data centers, whether it be over energy use, water consumption, zoning, and local permits, often assume that blocking data center construction would only affect the growth of AI. In reality, the same facilities that train AI models are often also used to process a small business’s payroll, host a school’s remote learning platform, or support a family’s long-distance FaceTime. Moreover, AI is increasingly being embedded into all of those online services to bring greater value to consumers.
Even a policy response that presumes to be narrowly tailored around AI risks disrupting the everyday digital services consumers, hospitals, banks, and small businesses depend on and for which we have no ready substitute.
This is especially important as data center energy consumption becomes a bigger part of the national conversation. The data center sector’s electricity use is rising, but the demand reflects the accumulated weight of everything Americans now do online, with AI adding to a base that streaming, banking, record-keeping, and e-commerce built over the past two decades. Although power use is rising, studies show that data centers shoulder a growing share of utilities’ fixed costs, reducing those costs that are allocated to residential electricity customers.
Digital infrastructure policy should be built around how Americans actually use the internet, not around whichever application is dominating the current news cycle. That means recognizing data centers as the shared foundation of the digital economy, which has been essential for communications, business, healthcare, banking, and entertainment – long before AI became a hot topic.