Yes, I watched Julia Roberts’ in the Erin Brockovich movie, and now she is back taking aim at the AI data center boom.
Environmental activist Erin Brockovich has introduced a crowdsourced mapping tool that helps communities track the rapid, resource-intensive expansion of AI data centers across the United States. The Brockovich AI Data Center Reporting website allows residents to log facilities that are operational, under construction, or rumored, and it organizes those entries into a single national view.
Of the more than 2,700 reports submitted so far, the largest concentration has come from Texas, where a 3-gigawatt project by MSB Global in Sulphur Springs has drawn substantial attention along with several active lawsuits tied to its roughly 1,600-acre site.
The map has reopened a familiar debate, and most of the conversation around it tends to settle on three things that are worth understanding before forming a view. Two of them give communities real reason for caution, and the third offers a genuine case in favor of the projects.
Why The AI Data Center Boom Worries Communities
The first concern is water. According to figures cited by the project and corroborated by Brookings and the Environmental and Energy Study Institute, a single large-scale AI data center can consume up to 5 million gallons of water a day, an amount comparable to the daily use of a town of as many as 50,000 people. That estimate represents the upper end of a wide range, and the actual draw of any given facility depends heavily on its cooling design, yet in regions already managing drought and limited freshwater, even the lower end of that range raises legitimate questions about long-term supply.
The second concern is electricity, and more specifically the question of who pays for it. The power requirements of these facilities are significant enough that utilities frequently need to upgrade transmission lines, substations, and related infrastructure to support them, and at least part of that cost commonly appears on the energy bills of ordinary residents who may never use the AI data center themselves. For scale, the largest technology companies, including Meta, Amazon, Microsoft, Google, and OpenAI, are expected to invest at least $700 billion in AI infrastructure this year, according to CNBC, which gives a sense of how quickly this demand is arriving.
The Economic Case For An AI Data Center
The third consideration, and the strongest argument in favor of these projects, is economic. For many of the communities involved, a large AI data center represents a substantial and durable source of revenue.
Sulphur Springs, per HeatMap News, offers a clear illustration, since the development there would generate an estimated $100 million in annual tax revenue, close to three times the city’s yearly budget, which helps explain why much of the most prominent litigation involves the city itself fighting to preserve the project. The industry has also begun addressing the environmental concerns directly.
Closed-loop cooling systems can reduce freshwater use by as much as 70 percent, Meta has said its new El Paso facility will use no water for most of the year and will aim to restore twice the volume it consumes to local watersheds, and Microsoft reports an 80 percent improvement in water efficiency across successive generations of its AI data center design. These approaches carry their own tradeoff, since reducing water use often requires additional electricity, which can shift some of the environmental burden toward carbon emissions depending on the local grid mix.
Taken together, these three considerations are part of what makes the map useful. By assigning a location, a water estimate, and an energy profile to facilities that often exist only as abstractions, the tool gives residents a concrete starting point for understanding what is being built near them. The project does come with an important caveat, since it is crowdsourced and includes rumored and proposed sites alongside confirmed ones, which means each entry works best as a signal that something merits a closer look before it is treated as confirmed.
The broader question the map raises is how communities, regulators, and companies will weigh the economic promise of AI infrastructure against its demands on water, power, and public trust.
As more of these projects move from announcement to construction, the organizations that treat water, energy, and community engagement as core design considerations, on par with chips and cooling, are likely to navigate the coming years more smoothly than those that overlook them.
You can contribute to the map by visiting the site, and the more useful exercise may be to ask who is at the table when the next AI data center is proposed in your own community.
