The AI Resource Grab: Why 230 Environmental Groups Want Congress to Pull the Plug

When you ask ChatGPT a question, you're not just tapping into artificial intelligence—you're also tapping into coal-fired power plants and draining local water supplies. That invisible cost is now triggering one of the most significant environmental backlashes the tech industry has faced in years.

In December 2025, a coalition of more than 230 environmental and community groups sent an urgent letter to Congress demanding something unprecedented: a federal moratorium on new data center construction. Led by Food & Water Watch and including heavyweights like Greenpeace and Friends of the Earth, the coalition argues that Big Tech is operating in what they call a "lawless wasteland," commandeering scarce resources to fuel an AI and cryptocurrency frenzy while ordinary families foot the bill.

"A lot of people don't see the benefits coming from AI and feel they will be paying for it with their energy bills and water," explains Emily Wurth of Food & Water Watch, capturing a growing sentiment that the AI revolution's costs are being socialized while its profits remain privatized.

The Zombie Power Plants

Perhaps the most striking accusation in the environmental coalition's case is that artificial intelligence is literally reversing decades of progress in the fight against climate change by breathing life back into coal plants that were supposed to die.

Across states like Georgia, Virginia, Nebraska, and Michigan, utilities are delaying the retirement of fossil fuel facilities specifically to feed the voracious appetite of data centers. In Georgia, Southern Company has announced plans to postpone shutting down the Plant Bowen and Plant Scherer coal facilities, pointing to load growth driven primarily by data centers. Virginia's Dominion Energy has made similar moves with the Clover Power Station, a coal-burning facility that was on the path to closure.

The numbers behind this reversal are staggering. Data center energy demand in the United States is projected to triple by 2030. Just one gigawatt of new data center capacity consumes enough electricity to power approximately 700,000 homes—and tech companies are racing to build dozens of these facilities.

These aren't the clean, ephemeral "clouds" of popular imagination. They're massive industrial complexes with an appetite for electricity that rivals heavy manufacturing, but with a crucial difference: unlike a steel mill or auto plant, their energy consumption is growing exponentially, not shrinking.

The Hidden Water Crisis

Energy consumption tells only half the story. AI's environmental footprint extends to another increasingly scarce resource: water.

The servers that power artificial intelligence generate enormous amounts of heat. Without constant cooling, they would literally melt down. That cooling requires water—vast quantities of it—which evaporates in the process and cannot be recovered.

Researchers at UC Riverside have calculated that a simple conversation with an AI chatbot, consisting of roughly 20 to 50 queries, consumes about 500 milliliters of water in cooling evaporation. That's a standard water bottle's worth of water, gone, for a brief digital interaction that most users assume is essentially free.

Scale that up to industrial levels, and the picture becomes alarming. A single large hyperscale data center can consume up to 5 million gallons of water per day—equivalent to the water usage of a town of 10,000 to 50,000 people. In Loudoun County, Virginia, which bills itself as the Data Center Capital of the World, water consumption by these facilities jumped 63 percent between 2019 and 2023.

This is happening at a time when water scarcity is becoming a defining challenge across much of the United States, from the drying Colorado River Basin to drought-stricken regions of the Southeast and Midwest.

The Economic Backlash

Environmental groups have strategically linked climate concerns with economic impacts, recognizing that wallet issues often mobilize broader coalitions than ecological arguments alone.

The fear is straightforward: ordinary ratepayers will subsidize the massive electrical grid upgrades required to support tech giants' data centers. In regions with high concentrations of these facilities, electricity bills have risen roughly 21 percent in recent years, outpacing inflation. Meanwhile, the companies consuming the power often negotiate special rates and tax incentives that shift costs onto residential customers.

Adding insult to injury, opponents argue that data centers represent "jobless growth." Once constructed, a massive facility that consumes resources equivalent to a heavy industrial factory might employ only 30 to 50 people, mostly in security and maintenance roles. The promised economic boom rarely materializes for local communities.

Communities Fight Back

The resistance isn't just coming from national environmental organizations. Local communities are increasingly saying no, and their opposition is having real impact. An estimated 64 billion dollars worth of data center projects have faced delays or outright cancellations due to local pushback.

Prince George's County in Maryland recently passed legislation limiting where data centers can be built. In Peculiar, Missouri—a town whose name has become somewhat ironic in this context—residents effectively blocked a 1.5 billion dollar project through zoning changes. Chesterton, Indiana saw a 1.3 billion dollar proposal withdrawn after sustained community outcry.

These aren't anti-technology Luddites; they're residents who looked at the trade-offs and decided the deal wasn't worth it. They saw water tables potentially threatened, electricity costs likely to rise, minimal job creation, and little tangible benefit to their daily lives.

The Standoff

The battle lines are now clearly drawn. Tech leaders like Eric Schmidt and Sam Altman frame data center expansion as essential to winning what they call the "AI arms race"—a geopolitical competition where America's technological dominance hangs in the balance. From their perspective, environmental concerns, while valid, must be balanced against national competitiveness and the transformative potential of artificial intelligence.

Environmental groups and affected communities counter that no technological advancement justifies backsliding on climate commitments or imposing hidden costs on the public. They argue that if AI is truly valuable, tech companies should pay the full cost of the resources they consume rather than externalizing those costs onto communities and the environment.

The demand for a federal moratorium represents a dramatic escalation in this conflict. Whether Congress will act remains uncertain, but the letter from 230 groups signals that AI's environmental reckoning has arrived. The question now is whether the technology sector will voluntarily adapt to address these concerns or whether it will take regulatory intervention to ensure that the digital future doesn't come at an unacceptable cost to the physical world we all inhabit.

As AI continues its rapid evolution, one thing is clear: the conversation about artificial intelligence can no longer ignore the very real resources—coal, water, and community wellbeing—being consumed in its creation.