By Jordan Watts
Knowledge is power, or so the saying goes, but in today’s digital economy of data storing all that knowledge takes power. A lot of power. Today, a single AI data center consumes enough electricity to keep the lights on in 80,000 U.S. households to be exact.
So where does all this power come from exactly? And, who is bankrolling all these data centers that keep popping up? Spoiler Alert: It might be you.
Data centers aren’t new, in fact they’ve been around for more than 50 years. Everything that you use and view online has to be stored somewhere; from your Spotify playlists to your email, “storage in The Cloud” doesn’t actually exist in the sky. Instead your data lives in giant warehouses full of servers around the world. These servers operate 24/7 and require energy intensive facilities capable of powering both the actual computers and massive cooling units necessary to keep them from overheating.
What’s new is the increased demand for data centers, requiring more and more power to keep up with the ongoing AI boom. Modern AI technology uses way more computing power than traditional internet searches, pulling from vast amounts of data and– you guessed it–power. Data centers in the United States are growing so rapidly that the Electric Power Research Institute predicts that they will account for 9% of all electricity demand in the country by 2030, double the amount today.
While data centers have always used lots of energy, AI data centers are particularly power hungry. One ChatGPT query can use more than 10 times as much power as the same search on Google, requiring not only more electricity but more water to keep servers from overheating as well.
Large data centers, like the ones necessary for AI computing, can require 100 megawatts (MW) or more per year according to the US Energy Information Administration. That’s equivalent to almost 400,000 electric vehicles, a medium sized power plant, or a small city.
It seems obvious that the tech giants building and utilizing these massive AI data centers should be the ones footing the bill, and they are– for the most part. As you would expect, the companies that build and use these AI data centers cover the cost of construction and power, but increased energy demand comes with its own costs.
In order to support the demand that these data centers put on the grid new infrastructure is required, something that utility customers normally foot the bill for. This would mean that homeowners, renters, and small businesses share the financial burden even if they don’t directly use the services. Seems unfair, right?
The thing is, attracting data centers is a no-brainer for utility companies which, at the end of the day, are still businesses looking to make a profit. Big projects that require significant revenue, like those required to support AI data centers, are huge moneymakers–regardless of who is footing the bill.
You'd be surprised to learn that while energy prices have been climbing for residential customers over the last five years, commercial prices have dipped. Largely because of TDSP charges, which make up a significant part of electricity costs and cover things like poles, wires, and transformers used to transmit powers to homes. This infrastructure must cover a wide area in order to reach folks, and it’s also the most vulnerable part of the grid to adverse and extreme weather. Commercial customers, like AI data centers, on the other hand, typically plug right into high-voltage lines, avoiding the distribution system entirely.
Residential customers also generally have the least lobbying power when it comes to negotiating pricing with electricity regulators. While tech giants have money and power to help them gain political allies, the average consumer doesn’t have a ton of influence when it comes to electricity pricing. This is part of what makes deregulated electricity markets, like Texas, so appealing with their competitive rates.
While AI data centers are popping up all over the country, Texas is either particularly well suited or vulnerable to them depending on how you look at things. The independent Texas electricity grid means that the state is especially appealing for tech companies looking to build and invest in AI data centers, thanks to competitive electricity rates and minimal regulations from the federal government.
The Energy Information Administration predicted that demand from large users (including, but not limited to, data centers) was predicted to grow 60% in 2025 and make up 10% of total demand on the Texas grid. Unlike other large consumers which have time sensitive energy needs, data centers require an uninterrupted supply of power and normally don’t participate in ERCOT’s high-demand response programs.
Stargate, a joint venture between the giants OpenAI, SoftBank, and Oracle plans to invest up to $500 billion in AI-related infrastructure primarily housed in the DFW area. The first of 30+ planned data centers is based in Abilene and will have 1.2 gigawatts (GW) of power capacity, which is equivalent to 1,200 megawatts (MW) when it’s fully completed later this year. That’s enough electricity to power more than a thousand homes.
When the entire facility is completed with dozens of data centers, the campus will have the potential to consume 10,512,000 MWh (megawatt hours) per year. For reference, that would be enough energy to power more than a million homes in 2025.
Unexpectedly, Google is also partaking in the data space race. In November of 2025 they announced a $40 billion investment to build two new data center campuses in Armstrong and Haskell for “new cloud and AI infrastructure.”
Unlike other tech giants though, Google has committed to “responsibly growing its infrastructure,” in part by putting money directly back into the Texas electricity grid. Alongside their investment in new campuses, the company announced a $30 million Energy Impact Fund to “scale and accelerate energy initiatives … and more than 6,200 MW of new energy generation.”
Social media conglomerate, Meta, has also taken up residency in the Lone Star State with an $1.5 billion dollar AI data center in El Paso. According to their announcement the campus will be almost 1.2 million square feet and “will have the ability to scale to 1 GW” of power capacity. As a reminder, that could power just under a thousand homes.
Since El Paso is not part of the deregulated electricity market in Texas, El Paso Electric was able to offer Meta a special electricity rate, making the location even more attractive. While the utility has said the rate difference would not be pushed onto residential consumers, it’s unclear whether or not this will hold true in the long run.
As of early 2026, Texas is home to more than 400 AI data centers operating or actively under construction and more continue to pop up. Significant clusters exist close to almost every major metropolitan area in the state, including DFW, Austin, Houston, San Antonio, and more.
Three main electricity grids power the United States and all of them are seeing an uptick in demand correlated with these new data centers. According to the main Texas grid operator, the demand for power in Texas is expected to nearly double by 2030. Continued demand like this will likely require large investments in transmission lines and generation to keep up.
While exact plans vary by state, many utilities and electricity regulators are negotiating with tech companies to balance energy demand growth and who will pay for it. Large infrastructure expansion for transmission and distribution lines are typically spread across all utility customers, but the average residential customer likely won’t need or use these upgrades.
To avoid burdening everyday consumers with increased costs, some states are exploring special tariffs and new customer rate classes for data centers. In practice this would help ensure that bigger customers cover more of the grid upgrade expenses. However, in order to meet utility revenue requirements to serve these growing electricity demands from data centers could still increase customer bills by 1% annually through 2032.
While it remains unclear exactly how much AI data centers will impact your electricity bills, if at all, it’s important to be prepared. You can protect yourself from rising electricity bills by comparing local energy plans and rates to ensure you get the best deal. To lock in the cheapest rate, EnergyBot recommends that you:
To find the best plan for your home or business, you can get started shopping today with EnergyBot.
