BY MICHELLE TALSMA EVERSON
Artificial intelligence may seem visible, but Arizona has none of its footprint. A network of sizable data centers are hidden beneath each AI search, chatbot swift, or image generator. These physical facilities run on land, energy, and frequently significant amounts of water. The focus is shifting from expanding to resources as Arizona continues to grow as a significant U.S. market for data centers.
According to some estimates, a single large data center can use up to two million liters ( more than half a million gallons ) of water per day, primarily for cooling. That figure is gaining more and more attention in a desert state where long-term water problems are currently being overcome.
Experts contend that the issue is more subtle than it might seem.
According to Chris Deaton, senior director of the Concerned Innovation Lab, “many people perceive AI as something intangible — a chatbot, an image generator, or a search tool.” However, a very natural infrastructure is hidden behind those interactions. AI is currently infrastructure. We can’t make the decision to do it without it.
The conversation is altered by the shift from software to infrastructure. Similar to electricity or telecommunications, AI requires long-term planning, resource management, and open accountability.
A WATER-STRESSED REGION AGREES IN A GROWING FOOTPRINT
Arizona’s appeal to developers of data centers is not accidental. The state offers land accessible, a fairly low risk of natural disasters, and a developing technology ecosystem supported by cloud computing and semiconductor manufacturing.
However, native growth has prompted questions. Residents of cities like Chandler have recently sparked open debate about proposed data centers, with residents raising concerns about water use and long-term sustainability.
Data centers are a significant portion of a far larger water landscape, according to Sarah Porter, director of the Arizona State University’s Kyl Center for Water Policy.
According to Porter,” Agriculture also uses more water than any other sector in spite of the region’s major urban growth and industrial development over the past few decades,” according to a new Kyl Center report examining large water users in northern Arizona.
The report argues that agriculture continues to be the main water user in the area despite the scrutiny that data centers are receiving. The report, which is titled” From Copper, Cattle and Cotton to Chips and Cloud Computing,” discovered that while comprehensive data on water use by data centers is still lacking, most facilities in central Arizona are served by municipal contractors with a Designation of Assured Water Supply ( DAWS ) from the Arizona Department of Water Resources. Providers must demonstrate they can meet projected water demand for decades before approving fresh development in accordance with this designation, helping protect existing users as fresh high-volume customers arrive online.
According to Porter,” The DAWS process forces providers to consider what kinds of water demand they want to take on” and protects existing water users from the effects of fresh high-volume users.
Given Arizona’s broader water problems, policymakers should give careful consideration to the possibility that massive water users may rely on unrenewable groundwater in some instances.
ENERGY IS NOT JUST WATER, BUT IT IS
Water is just a small component of the puzzle. Additionally, energy production for data centers can have its individual water requirements, depending on the source.
According to Porter,” Direct water consumption for data center power needs is also a crucial consideration.”
Although coming demand is questionable, Arizona utilities have increased water efficiency as they incorporate more renewable energy and natural gas.
This underscores the need for a more grounded conversation, in Deaton’s opinion.
The important issue is not whether these systems use resources. That’s a recognized given, he said. ” Everything infrastructure does. The real question is how appropriately they are created and run.
Concerned innovation frameworks encourage businesses to inquire about who benefits from new technologies, who may bear unexpected costs, and how risks are addressed before systems scale. They also encourage companies to ask this question.
That means shifting the emphasis away from treating AI as strictly online to decisions made in terms of transparency, planning, and engineering.
EXPECTATIONS AND TECHNOLOGY ARE EVOLVING, AND This ARE THEY.
Not all data centers use the exact amount of water. Water-intensive ambient cooling or closed-loop systems that reduce consumption are among the most prevalent cooling systems.
According to Porter,” We’ve found that data centers use a variety of cooling technologies.” ” Communities should demand that the water use plan used by the data centers they serve be compatible with the needs of the community.”
Some businesses are looking into alternatives to wet immersion cooling and air-cooled systems to reduce water consumption. Where possible, others use recovered or non-potable water.
Deaton stated,” There is absolutely progress happening. ” We’re seeing experimentation with AI-optimized energy efficiency, renewable-powered data centers, and zero-water cooling systems.”
These innovations are motivated by both cost and climate concerns. More cost-effective systems are frequently less costly to run over time. However, transparency varies.
Deaton said,” Communities frequently lack distinct visibility into resource use or long-term infrastructure planning.” He argued that the conversation should shift away from the topic of building and instead focus on how to build appropriately through collaboration between industry, policymakers, and communities.
POLICY TOOLS AND LOCAL CONTROL
Water policy decisions in Arizona are mostly made at the provincial level. Cities regulate water supply and set policies for big users, which allows for regional variations in approaches.
A new report from the Kyl Center revealed that 10 municipal water providers, who provide water to more than four million Arizonans, have passed ordinances or different laws to regulate large-volume water users. The second city in the US to pass a similar ordinance was Chandler.
These guidelines help to define what kind of expectations are there between residents and developers, including how well water is managed and what advantages do projects bring.
According to Porter,” they live up to the community’s expectations in terms of the return on investment from a high-volume user.”
That means that residents can influence how projects develop.
WHAT APPEARS IN THE FUTURE
The infrastructure supporting AI will continue to grow as well. The challenge for Arizona is finding a balance between economic constraints and economic opportunity.
According to Deaton, the emphasis should be on guiding innovation appropriately rather than slowing it down.
Responsible growth refers to recognizing two realities at once, such as that the Southwest has actual environmental constraints and that AI infrastructure is becoming increasingly important financially.
That means enhancing efficiency, enhancing transparency, and promoting long-term planning for water and energy use.
Deaton said,” The goal isn’t to stifle innovation.” It’s important that the systems we create today are as useful as we navigate this transition.
That balance will affect Arizona’s ability to determine both the sustainability of the communities that support it and the future of technology.
