The lightning-fast evolution of Artificial Intelligence (AI) is commanding our attention, sparking conversations about breakthroughs in healthcare, solutions for climate change, and the very fabric of our digital lives. Yet, beneath the surface of these exciting advancements lies a less glamorous, and often contentious, reality: the immense and growing hunger for energy that AI demands. In a surprising twist, this insatiable appetite is giving a second life to an industry that has long been a focal point of environmental concerns – fracking.
The Unseen Energy Drain: AI’s Thirst for Power
As AI companies race to build the most powerful algorithms and develop groundbreaking applications, they are also constructing colossal data centers. These digital fortresses, housing the powerful chips and vast storage needed to train and run sophisticated AI models, require an enormous amount of electricity. And in a trend that has flown largely under the radar, many of these AI powerhouses are increasingly tapping directly into fossil fuels, particularly natural gas extracted through hydraulic fracturing, or fracking.
This isn’t a theoretical concern; it’s a rapidly unfolding reality. AI firms are strategically locating their massive data centers near major natural gas production sites, often establishing their own power generation capabilities by directly accessing these fossil fuel resources. This approach, while seemingly pragmatic for meeting immediate energy needs, raises significant ethical and environmental questions for the communities that find themselves in the shadow of these energy-intensive operations.
A Texas-Sized Footprint: Poolside’s Ambitious Data Center
A prime example of this trend is the recent announcement of Poolside, an AI coding assistant startup. The Wall Street Journal reported that Poolside is developing a sprawling data center complex in West Texas, a region synonymous with oil and gas extraction. Covering over 500 acres – a footprint two-thirds the size of New York’s Central Park – this facility, codenamed Horizon, is set to become a behemoth of computing power. Its defining feature? It will generate its own electricity by tapping directly into the abundant natural gas reserves of the Permian Basin, the nation’s most prolific oil and gas field. Fracking, as the primary extraction method in this region, is not just common; it’s the very engine driving this AI energy project.
The sheer scale of Horizon is staggering. It will produce a colossal two gigawatts of computing power. To put that into perspective, that’s equivalent to the entire electric generating capacity of the Hoover Dam. However, instead of harnessing the flow of the Colorado River, Horizon will be powered by the combustion of fracked natural gas.
Poolside is collaborating with CoreWeave, a specialized cloud computing company that provides access to an extensive array of Nvidia AI chips. CoreWeave will be supplying Horizon with access to over 40,000 of these cutting-edge processors, underscoring the advanced AI capabilities this facility is designed to support.
The Wall Street Journal aptly described this burgeoning landscape as an “energy Wild West,” a term that encapsulates the rapid, often unregulated, expansion of AI infrastructure in regions rich with fossil fuels.
Beyond Poolside: A Widespread Phenomenon
Poolside’s ambitious project is far from an isolated incident. The pursuit of abundant and readily available energy has led many of the major players in the AI arena to adopt similar strategies. OpenAI, a pioneer in generative AI, is another prominent example. CEO Sam Altman recently toured the company’s flagship Stargate data center in Abilene, Texas – a location conveniently situated within range of the Permian Basin’s natural gas supply.
Altman’s candid admission during the tour, “We’re burning gas to run this data center,” highlights the direct reliance on fossil fuels. The Stargate complex is an impressive undertaking, requiring approximately 900 megawatts of electricity across eight buildings. It even includes a new gas-fired power plant, equipped with turbines comparable to those found on naval warships, according to reports from the Associated Press.
While the companies involved assert that this on-site power plant primarily serves as a backup, with the bulk of electricity drawn from the local grid, the grid itself is a complex mix. It draws power from natural gas sources and the vast wind and solar farms that dot the West Texas landscape. However, for the residents living in close proximity to these facilities, this energy mix offers little comfort.
The Human Cost: Disruption and Disquiet in Host Communities
Arlene Mendler, a resident living directly across the street from the Stargate data center, shared her frustrations with the Associated Press. She expressed a deep sense of disenfranchisement, noting that no one had sought her opinion before massive tracts of mesquite shrubland were bulldozed to make way for the facility. For decades, Mendler sought the peace and quiet of rural Texas, but now, the constant hum of construction and the glare of bright lights have irrevocably altered her environment and spoiled her nighttime views.
Beyond the immediate disruption, the implications for water resources in drought-prone West Texas are a significant source of anxiety for local communities. During Altman’s visit, the city’s reservoirs were reportedly at only half capacity, and residents were already subject to strict twice-weekly outdoor watering schedules.
While Oracle, a technology giant involved in data center infrastructure, claims that each of its data center buildings will require a modest 12,000 gallons per year for cooling systems after an initial fill, a professor from the University of California, Riverside, who specializes in AI’s environmental impact, argues this assessment is misleading. These sophisticated cooling systems, he points out, are more energy-intensive. This increased energy demand translates to a higher indirect water consumption at the power plants that generate the electricity, exacerbating water scarcity concerns in already parched regions.
Meta’s Massive Undertaking and Local Resistance
Meta, the parent company of Facebook and Instagram, is also aggressively expanding its data center footprint, with a similar reliance on fossil fuels. In Richland Parish, Louisiana – one of the state’s poorest regions – Meta plans to construct a colossal data center. This facility, the equivalent in size of 1,700 football fields, will demand an astounding two gigawatts of power solely for its computational needs.
To meet this demand, the local utility company, Entergy, is investing a substantial $3.2 billion to build three large natural gas power plants. These plants, with a combined capacity of 2.3 gigawatts, will primarily burn natural gas extracted through fracking in the nearby Haynesville Shale formation. Similar to residents in Abilene, the people of Louisiana are expressing dismay at the prospect of round-the-clock construction and the increased environmental burden.
It’s worth noting that Meta is also developing a data center in El Paso, Texas, near the New Mexico border. This facility, projected to have one gigawatt of capacity, is expected to come online in 2028. Meta has stated that this El Paso project will be powered by 100% clean and renewable energy sources, a positive development that stands in contrast to its Louisiana venture.
Even Elon Musk’s xAI Has Fracking Ties
The reach of this trend extends even to Elon Musk’s xAI. The company’s facility in Memphis, Tennessee, has been the subject of considerable controversy this year. While Memphis Light, Gas and Water (MLGW) currently supplies power to xAI, the company is in the process of building its own substations. MLGW procures natural gas on the spot market, and it is transported to Memphis via two major pipeline companies: Texas Gas Transmission Corp. and Trunkline Gas Company. Both of these suppliers rely on natural gas sourced from hydraulically fractured shale formations, highlighting a direct link between xAI’s operations and the fracking industry.
The Geopolitical Imperative: Battling China for AI Supremacy?
When questioned about the rationale behind these energy choices, AI companies often cite a compelling geopolitical argument: the need to outpace China in the global AI race. Chris Lehane, OpenAI’s Vice President of Global Affairs, articulated this perspective during a recent interview. He emphasized that in the near future, the United States will likely require an astonishing "gigawatt of energy a week" to support its AI ambitions.
Lehane pointed to China’s massive energy infrastructure buildout, including the construction of 450 gigawatts of capacity and 33 nuclear facilities in the past year alone, as a benchmark for the scale of investment required. He framed the decision to locate data centers in economically challenged areas like Abilene or Lordstown, Ohio, as part of a broader strategy to re-industrialize the country, bring manufacturing back, and modernize energy systems.
Government Support: A Boost for Fossil Fuels
The Trump administration has shown a clear inclination to support this direction. An executive order signed in July 2025 is designed to accelerate the development of gas-powered AI data centers by streamlining environmental permitting processes, offering financial incentives, and opening federal lands for projects utilizing natural gas, coal, or nuclear power. Notably, this executive order explicitly excludes renewable energy sources from its support mechanisms.
The Unseen Carbon Footprint: Consumer Ignorance and Corporate Strategy
For the vast majority of AI users, the underlying energy consumption and carbon footprint of their digital tools remain largely invisible. Their focus is often on the dazzling capabilities of AI, such as OpenAI’s Sora 2, a hyperrealistic video generation product that demands exponentially more energy than a simple chatbot. The source of the electricity powering these innovations is a secondary concern, if it’s a concern at all.
AI companies, for their part, have strategically positioned natural gas as the pragmatic and, in their view, inevitable solution to the escalating power demands of AI. This narrative, however, risks overlooking the significant environmental consequences and the long-term implications of a massive fossil fuel buildout.
A House of Cards? The Interdependencies of the AI Ecosystem
The AI sector has become a complex web of mutual dependencies. OpenAI relies on Microsoft for cloud infrastructure and funding. Microsoft, in turn, depends on Nvidia for its crucial AI chips. Nvidia, alongside companies like Broadcom, is a key supplier for chip manufacturers. Oracle provides essential database and cloud services, and data center operators are the physical backbone for all of it. This creates a self-reinforcing loop where each entity buys from and sells to the others.
The Financial Times has highlighted the potential fragility of this ecosystem. If a foundational element cracks, it could leave behind a significant amount of expensive, underutilized infrastructure – both digital and fossil-fuel dependent.
Is All This New Capacity Necessary? A Question of Efficiency
A critical question that has been largely absent from the public discourse is whether the sheer volume of new energy-intensive data center capacity being built is truly necessary. Research from Duke University suggests there might be a more efficient path forward. Their findings indicate that utility companies typically operate at only about 53% of their available capacity throughout the year. This suggests considerable latent capacity that could potentially accommodate new demand without necessitating the construction of entirely new power plants.
According to estimates from MIT Technology Review, if data centers were to implement measures to reduce electricity consumption by roughly half for just a few hours during the annual peak demand periods, utilities could potentially handle an additional 76 gigawatts of new load. This would be more than enough to absorb the projected 65 gigawatts of demand from data centers by 2029.
Such flexibility would not only allow for faster deployment of AI data centers but, more importantly, could provide a crucial window for utilities to develop cleaner, renewable energy alternatives. This approach would offer a reprieve from the current rush to build new natural gas infrastructure.
The Long Shadow of Fossil Fuels: Who Bears the Cost?
However, the prevailing narrative, driven by the urgency to compete with autocratic regimes like China, suggests that this path of building new natural gas infrastructure is likely to continue. This trajectory risks burdening regions with more fossil-fuel dependent power plants and leaving residents with escalating electricity bills to finance the initial investments. These financial obligations could persist for years, even after the tech companies’ energy contracts expire.
Meta, for instance, has committed to covering Entergy’s costs for the new Louisiana generation for 15 years. Poolside’s lease with CoreWeave also spans 15 years. The long-term implications for customers when these contracts end remain a significant unanswered question.
A Glimmer of Hope: The Rise of Clean Energy Alternatives
Despite the current reliance on fossil fuels, there are promising developments indicating a potential shift towards cleaner energy sources. Significant private investment is being channeled into the development of small modular reactors (SMRs) and advanced solar installations, with the expectation that these cleaner alternatives will eventually become the primary energy providers for data centers.
Fusion energy startups, such as Helion and Commonwealth Fusion Systems, have also attracted substantial funding from key players on the front lines of AI, including Nvidia and OpenAI’s Sam Altman. This optimism is not confined to private investment circles; it has permeated public markets, with several speculative energy companies going public based on the anticipation that they will one day power the data centers of the future.
The Unasked Question: What About the People?
In the interim, which could still span decades, the most pressing concern is the disproportionate burden placed on the people who inhabit these regions. They are the ones who will ultimately bear the financial and environmental costs of this rapid energy buildout, often without having been consulted or having had a say in the matter. As AI continues its relentless advance, the energy choices made today will have profound and lasting consequences for communities and the planet.
Key Takeaways for a Balanced Perspective:
- AI’s Energy Demand is Skyrocketing: The growth of AI necessitates massive amounts of electricity, driving the construction of large-scale data centers.
- Fracking’s Resurgence: Many AI companies are opting to power these data centers by tapping into natural gas extracted through fracking, particularly in regions like West Texas and Louisiana.
- Community Impacts: Residents in host communities face disruptions, environmental concerns, and potential water shortages due to these projects.
- Geopolitical Motivations: The drive to compete with China is a significant factor influencing AI companies’ energy strategies.
- Government Incentives: Policy decisions are actively favoring fossil fuel-based energy solutions for AI infrastructure.
- Efficiency and Alternatives: There’s a debate about whether existing energy infrastructure could be used more efficiently and the potential of emerging clean energy technologies.
- Long-Term Costs: The financial and environmental burdens of new fossil fuel infrastructure may fall disproportionately on local communities.
As AI continues to integrate into every facet of our lives, understanding the intricate energy demands and the choices being made today is crucial for shaping a sustainable and equitable future.
Leave a Reply