Your AI tools work on fracked gas and bulldozed Texas lands

The era of artificial intelligence is giving fracking a second act, a surprising development for an industry that, even during its boom years in the early 2010s, climate advocates blamed for poisoning the water table, man-made earthquakes, and the stubborn persistence of fossil fuels.

AI companies are building massive data centers near major gas production sites, often generating their own energy by tapping directly into fossil fuels. It’s a trend that’s been overshadowed by headlines about the intersection of AI and healthcare (and the climate change solution), but one that could reshape — and raise difficult questions for — the communities that host these facilities.

Take the most recent example. The Wall Street Journal reported this week that AI programming assistant startup Poolside is building a data center park on more than 500 acres in West Texas — about 300 miles west of Dallas — an area two-thirds the size of Central Park. The facility will generate its power by tapping natural gas from the Permian Basin, the country’s most productive oil and gas field, where hydraulic fracturing is not only popular, it’s the only game in town.

The project, called Horizon, will produce 2 gigawatts of computing power. This is equivalent to the entire electrical capacity of the Hoover Dam, except that instead of tapping the Colorado River, it burns fracked gas. Poolside is developing the facility with CoreWeave, a cloud computing company that leases access to Nvidia AI chips and which provides access to more than 40,000 of them. The magazine calls it “Wild West energy,” which seems appropriate.

However, Poolside is not alone. Almost all major players in the field of artificial intelligence are following similar strategies. Last month, OpenAI CEO Sam Altman toured his company’s flagship Stargate data center in Abilene, Texas — about 200 miles from the Permian Basin — where he was frank: “We’re burning gas to run this data center.”

The complex requires about 900 megawatts of electricity across eight buildings, and includes a new gas-fired power plant that uses turbines similar to those that power warships, according to the Associated Press. The companies say the station only provides backup power, with most of the electricity coming from the local grid. This grid, for the record, is drawn from a combination of natural gas and sprawling wind and solar farms in West Texas.

But people living near these projects do not feel completely comfortable. Arlene Mendler lives across the street from Stargate. She told the AP she wished someone would ask her opinion before bulldozers cleared a large swath of mesquite brush to make way for what was being built on top of it.

“It completely changed the way we lived,” Mendler told the AP. She moved to the area 33 years ago in search of “peace, quiet and serenity.” Now the construction was the background soundtrack, and the bright lights of the scene spoiled her night views.

Then there is the water. In drought-prone West Texas, local residents are particularly concerned about how new data centers will impact water supplies. The city’s reservoirs were at about half capacity during Altman’s visit, and residents were on a twice-weekly outdoor watering schedule. Oracle claims each of the eight buildings will need only 12,000 gallons per year after the initial 1-million-gallon fill of closed-loop cooling systems. But Shaoli Ren, a professor at the University of California, Riverside who studies the environmental footprint of artificial intelligence, told the AP that this is misleading. These systems require more electricity, which means more indirect water consumption in the power plants that generate that electricity.

Meta follows a similar strategy. In Richland Parish, Louisiana’s poorest area, the company plans to build a $10 billion data center the size of 1,700 football fields that will require 2 gigawatts of power for computing alone. Utility company Entergy will spend $3.2 billion to build three large, 2.3-gigawatt natural gas power plants to feed the facility by burning gas extracted through hydraulic fracturing in the nearby Haynesville Shale. Louisiana residents, like Abilene residents, are not thrilled about being surrounded by bulldozers around the clock.

(Meta is also building in Texas, but elsewhere in the state. The company this week announced construction of a $1.5 billion data center in El Paso, near the New Mexico border, with one gigawatt of capacity expected to be up and running in 2028. El Paso is not near the Permian Basin, and Meta says the facility will be powered by 100% clean, renewable energy. One point for Meta.)

Even Elon Musk’s xAI, whose Memphis facility sparked controversy this year, has connections to fracking. Memphis Light, Gas and Water — which currently sells power to xAI but will eventually own the substations that xAI builds — buys natural gas on the spot market and transports it to Memphis through two companies: Texas Gas Transmission Corp. and Trunkline Gas Company.

The Texas Gas Transmission is a two-way pipeline carrying natural gas from Gulf Coast supply areas and several major hydraulically fractured shale formations through Arkansas, Mississippi, Kentucky and Tennessee. Trunkline Gas, another Memphis supplier, also transports natural gas from hydraulic fracturing sources.

If you’re wondering why AI companies are following this path, they’ll tell you it’s not just about electricity; It’s also about beating China.

That was the argument made by Chris Lehane last week. Lehane, a veteran political activist who joined OpenAI as vice president of global affairs in 2024, laid out the case during an on-stage interview with TechCrunch.

“We believe that in the not too distant future, at least in the United States, and around the world, we will need to generate close to a gigawatt of power per week,” Lehane said. He pointed to China’s huge energy reserve: 450 gigawatts and 33 nuclear facilities were built last year alone.

When TechCrunch asked about Stargate’s decision to build in economically challenged areas like Abilene or Lordstown, Ohio, where more gas-fired plants are planned, Lehane returned to geopolitics. “So we [as a country] If you do it right, you have an opportunity to reindustrialize countries, reindustrialize, and also transform our energy systems so that we can have the modernization that needs to happen.

The Trump administration is certainly on board. The July 2025 executive order accelerates gas-fired AI data centers by streamlining environmental permits, offering financial incentives, and opening federal lands to projects that use natural gas, coal, or nuclear power — while explicitly excluding renewables from support.

Right now, most AI users remain largely unaware of the carbon footprint behind their fancy new toys and work tools. They’re focusing more on capabilities like Sora 2 — OpenAI’s hyper-realistic video generation product that requires significantly more power than a simple chatbot — rather than on where the electricity comes from.

Companies are counting on this. They position natural gas as a practical and inevitable answer to the increasing energy demands of artificial intelligence. But the speed and scale of fossil fuel construction deserves more attention than it gets.

If this were a bubble, it wouldn’t be pretty. The AI ​​sector has become a circular firing squad of dependencies: OpenAI needs Microsoft, Nvidia needs Broadcom, and Oracle needs data center operators who need OpenAI. They all buy and sell to each other in a self-reinforcing loop. The Financial Times noted this week that if the foundation cracks, there will be a lot of expensive infrastructure left standing, both the digital and the gas-fired kind.

OpenAI’s ability alone to meet its obligations “is a growing concern for the broader economy,” the outlet wrote.

One key question that has been largely absent from the conversation is whether or not all these new capabilities are necessary. A Duke University study found that utilities typically use only 53% of their available energy throughout the year. This suggests there is plenty of room to absorb new demand without building new power plants, as MIT Technology Review reported earlier this year.

Researchers at Duke University estimate that if data centers cut electricity consumption by about half for just a few hours during annual peak demand periods, the facilities could handle an additional 76 gigawatts of new load. This would effectively accommodate the 65 GW data centers are expected to need by 2029.

This kind of flexibility would allow companies to launch AI data centers faster. More importantly, it could provide a respite from the rush to build natural gas infrastructure, giving utilities time to develop cleaner alternatives.

But again, that would mean losing ground to an authoritarian regime, according to Lehane and many others in the industry, so instead, a natural gas building spree seems likely to burden regions with more fossil fuel plants and leave residents with higher electricity bills to finance the day’s investments, including long after the technology companies’ contracts expire.

For example, Meta guaranteed it would cover Entergy’s costs for new Louisiana generation for 15 years. Poolside’s lease with CoreWeave runs for 15 years. What happens to customers when these contracts expire remains an open question.

Things may change eventually. Much of the private money is being directed to small modular reactors and solar installations with the expectation that these clean energy alternatives will become more centralized power sources for these data centers. Likewise, fusion startups like Helion and Commonwealth Fusion Systems have raised significant funding from AI frontliners, including Nvidia and Altman.

This optimism is not limited to private investment circles. This excitement has spilled over into public markets, where many of the “non-revenue” energy companies that have managed to go public have truly forward-thinking market valuations, based on the expectation that they will one day power these data centers.

In the meantime — which could last for decades — the more pressing concern is that the people they’ll be leaving holding the bag, both financially and environmentally, never asked for any of this in the first place.