The artificial intelligence (AI) era is unexpectedly fueling a significant resurgence in the fracking industry, a surprising turn for a sector previously criticized for environmental damage, including poisoned water tables and man-made earthquakes. While public attention often focuses on AI's applications in healthcare or climate solutions, the construction of massive AI data centers near major gas-production sites, often generating their own power by tapping directly into fossil fuels, is a significant, yet often overlooked, trend. This development raises critical questions for communities hosting these facilities and could reshape industrial landscapes, bringing difficult challenges.

A notable example surfaced recently with reports that AI coding assistant startup Poolside is developing a data center complex on over 500 acres in West Texas, an area two-thirds the size of Central Park. This facility, dubbed Horizon, will generate its own 2 gigawatts of computing power—equivalent to the Hoover Dam’s entire electric capacity—by burning natural gas from the Permian Basin, the nation’s most prolific oil and gas field where hydraulic fracturing is the primary method of extraction. Poolside is collaborating with CoreWeave, a cloud computing company providing access to more than 40,000 Nvidia AI chips for the project, which has been aptly described by the Wall Street Journal as an “energy Wild West.”

Poolside is not alone in this strategy. Nearly all major AI players are pursuing similar paths. OpenAI CEO Sam Altman, for instance, visited his company’s flagship Stargate data center in Abilene, Texas, approximately 200 miles from the Permian Basin, where he candidly acknowledged, “We’re burning gas to run this data center.” This complex requires about 900 megawatts of electricity across eight buildings and incorporates a new gas-fired power plant utilizing turbines similar to those found in warships. While the companies claim this plant primarily provides backup power, with most electricity coming from the local grid—which draws from a mix of natural gas and West Texas wind and solar farms—local residents remain unconvinced and concerned.

Residents near these projects express significant unease. Arlene Mendler, living across the street from Stargate, lamented the bulldozing of a vast tract of mesquite shrubland to make way for the facility, stating it “has completely changed the way we were living” and disrupted her previously peaceful existence with constant construction noise and bright nighttime lights. Water scarcity is another major concern in drought-prone West Texas, especially as city reservoirs were at roughly half capacity during Altman’s visit, with residents on a twice-weekly outdoor watering schedule. While Oracle, involved in the Stargate project, asserts minimal water use (12,000 gallons per building annually after an initial million-gallon fill for closed-loop cooling), a University of California, Riverside professor, Shaolei Ren, who studies AI’s environmental footprint, argues this is misleading, as increased electricity demand translates to higher indirect water consumption at power plants generating that electricity.

Meta is also adopting a similar strategy. In Richland Parish, Louisiana, one of the state's poorest regions, Meta plans a $10 billion data center the size of 1,700 football fields, requiring 2 gigawatts of computational power alone. To supply this demand, utility company Entergy will invest $3.2 billion to construct three large natural-gas power plants with 2.3 gigawatts of capacity, burning gas extracted via fracking from the nearby Haynesville Shale. Like their Texan counterparts, Louisiana residents are not pleased with round-the-clock bulldozing and construction. Notably, Meta is also building a $1.5 billion data center in El Paso, Texas, near the New Mexico border, with 1 gigawatt capacity expected online in 2028, stating this facility will be matched with 100% clean and renewable energy, differentiating it from some other projects.

Even Elon Musk’s xAI facility in Memphis has indirect fracking connections. Memphis Light, Gas and Water (MLGW), which currently sells power to xAI and will eventually own the substations xAI is building, procures natural gas on the spot market. This gas is piped to Memphis through Texas Gas Transmission Corp. and Trunkline Gas Company, both of which carry natural gas from hydraulically fractured shale formations, thus linking xAI to the fracking supply chain.

AI companies justify this reliance on fossil fuels not only for electricity but also as a strategic imperative against geopolitical rivals like China. Chris Lehane, OpenAI’s vice president of global affairs, articulated this perspective during an interview, forecasting a need for approximately “a gigawatt of energy a week” in the U.S. and globally. He highlighted China’s massive energy expansion, including 450 gigawatts and 33 nuclear facilities constructed in the last year, underscoring the competitive drive.

Lehane further connected the choice of economically challenged locations, such as Abilene, Texas, or Lordstown, Ohio—where more gas-powered plants are planned—to a broader vision. He suggested that “If we [as a country] do this right, you have an opportunity to re-industrialize countries, bring manufacturing back and also transition our energy systems so that we do the modernization that needs to take place.” This narrative positions the fossil fuel buildout as a means to both national security and economic revitalization.

The Trump administration has actively supported this direction. A July 2025 executive order is set to fast-track gas-powered AI data centers by streamlining environmental permits, offering financial incentives, and opening federal lands for projects using natural gas, coal, or nuclear power, while pointedly excluding renewables from similar support.

Currently, most AI users remain largely unaware of the carbon footprint behind their advanced tools and work tools, focusing more on capabilities like OpenAI’s energy-intensive Sora 2 video generation product—which requires exponentially more energy than a simple chatbot—than on energy sourcing. AI companies are leveraging this unawareness, portraying natural gas as the unavoidable and pragmatic solution to AI’s burgeoning power demands. The AI sector itself has become a complex web of dependencies, with companies like OpenAI, Microsoft, Nvidia, Broadcom, and Oracle forming a self-reinforcing economic loop. The Financial Times has warned of potential instability if this foundation cracks, leaving behind expensive digital and gas-burning infrastructure, and raising concerns about OpenAI’s obligations for the wider economy.

A crucial aspect often missing from the discussion is whether this extensive new capacity is genuinely necessary. A Duke University study suggests that utilities typically use only 53% of their available capacity annually, implying substantial room to absorb new demand without constructing additional power plants. The researchers estimate that if data centers reduced electricity consumption by roughly half for a few hours during annual peak demand periods, utilities could manage an additional 76 gigawatts of new load, effectively accommodating the 65 gigawatts projected for data centers by 2029. Such flexibility could accelerate AI data center launches and, more critically, offer a reprieve from the rush to build natural gas infrastructure, allowing time for cleaner alternatives to develop.

However, the industry's geopolitical focus (e.g., Lehane's argument about losing ground to autocratic regimes) appears to override such considerations. Consequently, the natural gas building spree is likely to burden regions with more fossil-fuel plants and leave residents facing soaring electricity bills to finance these current investments, potentially long after the tech companies’ contracts expire. For instance, Meta has guaranteed to cover Entergy’s costs for Louisiana's new generation for 15 years, and Poolside’s lease with CoreWeave also runs for 15 years, leaving the long-term impact on customers uncertain.

Despite the immediate fossil fuel reliance, private investment is significantly flowing into cleaner energy alternatives, such as small modular reactors (SMRs) and solar installations, with expectations that they will eventually become central energy sources for data centers. Fusion startups like Helion and Commonwealth Fusion Systems have also attracted substantial funding from key AI figures, including Nvidia and Sam Altman. This optimism extends to public markets, where “non-revenue-generating” energy companies with truly anticipatory market caps are betting on future roles in powering data centers.

In the interim, which could span decades, the most pressing concern remains that the local communities, who stand to bear the financial and environmental costs, never consented to this large-scale industrial transformation. The rapid and massive fossil fuel buildout to power AI demands more scrutiny and public awareness than it is currently receiving, highlighting the potential for significant, long-term consequences if this energy bubble proves unsustainable.