How AI data centers are quiet competitors in the housing market

Transcript:

Allie Barefoot: Welcome back to Data in Context. I'm Allie Barefoot with Cotality. There's a quiet competitor in the housing market that doesn't necessarily want your dream home, but the very land that it's built on. The AI housing boom started in San Francisco, where homes are being snapped up by high-paying AI employees. Now, major data centers are competing with residential developers, not just for land, but for power. So what happens when the cloud literally hits the ground?

Well, today we're going to be discussing how data centers affect residential projects with Cotality's Head of Data Science, Amy Gromowski, to determine if the housing market is being outpaced by its new neighbors. Let's put the data into context. Welcome in to Data in Context, Amy.

Amy Gromowski: Thanks for having me, Allie. Excited to talk to you today.

Allie Barefoot: Yeah, it's a very exciting topic, and you've seen the evolution of AI for the last 25 years, so you're a great person to talk to about this. Because we used to discuss AI in the terms of a cloud, like it was an ethereal thing. But now that cloud has an actual physical address, and it's moving into our neighborhoods. So how is the AI boom fundamentally changing American real estate?

Amy Gromowski: It's true. AI infrastructure is bringing the boom to real estate. It's in the same category as railroads, as oil, but really with one big distinction: railroads moved goods, AI infrastructure is moving intelligence. So, yes, the cloud—it did once feel like it lived in the atmosphere, but it's actually always been on land. It's just that now that land is in our neighborhoods. So, you know, because AI proximity requires—proximity to power, fiber, people—it's really fundamentally changing real estate by turning the pockets of America that are very utility-rich into the new gold mines, regardless of their traditional, you know, prestige or location.

Allie Barefoot: That's a super interesting way to think about that. Railroads, they transport goods, but AI is transporting intelligence. And because they're in that same category in terms of how they affect the real estate market, how does AI reshape where and how people live, and how they find homes?

Amy Gromowski: Yeah, so to use the railroad analogy, railroads created hub cities. So think Chicago; Omaha, Nebraska, right? Even though that, compared to Chicago, can feel pretty rural, it absolutely is a hub city. And proximity to a rail line really skyrocketed land value. In the same way, you know, when oil struck on land, that skyrocketed land value, and oil rigs created boom towns. So now with the development drivers for AI, it is shifting that growth to the rural areas. So approximately 67% of all planned data centers are located in rural areas where power grids have more capacity and land is readily available. But that velocity of change is different. And I just find this so interesting, right? The railroad took 50 years to reshape the U.S. economy. The AI buildout is happening in a single decade. In 2025, the tech giants invested over 320 billion in infrastructure, and that's nearly 5% of the U.S. GDP at its peak. So that matches the historical scale of the railroad boom, but at five times the speed.

Allie Barefoot: Right, and you can almost feel that speed. You know, AI's been around for a while, but it feels like recently it's getting a surge in conversations, in our neighborhoods, in our data. And in our most recent survey here at Cotality that we did in Q1, we found that people are actually trusting AI less. We found a 14-point drop in trust from just last year. Once again, it's just constantly evolving. And this seems to go against the narrative that the more automation is better. So what do you think it'll take for people to have more trust in AI and, therefore, make us all—make all this power being built make sense?

Amy Gromowski: Yeah, I found that stat really interesting, too, that it dropped from 30% in 2025 to 16%. I've been talking about that quite a bit. It really comes down to: buyers aren't rejecting the efficiency of AI, they are rejecting its opacity, right? So the study also showed that 68% of buyers want to manually verify that AI-provided housing information, and 64% are worried that AI is just recycling unverified data, which makes me feel really good about where we're at for Cotality in terms of this idea that we can provide certainty as the product itself, right? We have so much in terms of the validation of our data pipelines and in the diversity of sources. And I can go on and on there, right? In terms of how we can provide certainty of product. But really what it comes down to is, you know, 44% of buyers in this survey would actually pay an extra fee to have a human expert verify an AI's housing decisions. So it's about trust, and that trust requires transparency. And buyers need to know when AI is making a decision, and they want to know what data it is using.

Allie Barefoot: And that makes complete sense because, you know, when I think of AI, I'm thinking of computer codes, I'm thinking of this big thing that's looming above us, but it's actually happening right here, right now. And, for example, we're kind of seeing it in San Francisco—you know, they're leading the AI boom—and Abilene is home of projects like Stargate, and they're creating different narratives here around what AI means in the housing sphere. And when San Fran gets the software wealth, and Abilene gets the physical burden of building infrastructure, do you think that this friction where demand for AI grows faster than our ability to build for it is just a temporary bottleneck, or is this a new normal for the housing market itself?

Amy Gromowski: I think we're looking at a new normal. There's a real shift where capital allocation is heavily skewing toward the digital infrastructure. And I think this dynamic that you're raising is really interesting. So let's look at the San Francisco software wealth. San Francisco has now re-established itself as an undisputed ground zero for the global AI boom. So the Bay Area has captured approximately 80% of the U.S. national startup funding in Q1 here of 2026. And that influx of AI wealth has pushed San Francisco's median home price to more than 1 and a half million. That's where Cotality's AVMs say, you know, the median single-family home prices are, right? And that's for March of 2026. That's up 13% from a year ago. So that's definitely a wealth boom.

But then you look at where the data centers are going, and data centers are arguably the most capital-dense real estate asset—more than anything else, right, on the planet. And a single campus can represent billions of dollars in capital investment. So, let's take Project Stargate in Abilene, Texas. It's a 500 billion, excuse me, artificial intelligence infrastructure initiative, and construction is underway on a 1200-acre campus. So for towns like Abilene, from a local government's perspective, the data centers are incredibly lucrative. They generate a massive amount of property and equipment taxes. But unlike building a new residential subdivision, they don't add hundreds of children to the local school system, they don't require a massive increase in local police or fire services, right, for example. But that tax revenue is also able to fund for existing communities' schools, parks, public services, while simultaneously, the property taxes from a personal perspective are actually lowered for the residents. So if you think about these two—the San Francisco wealth and the urban boom, right, with data centers—these are just two very different dynamics from a real estate perspective.

Allie Barefoot: And let's stay in Abilene for a minute here because Cotality data shows that rents in that city are skyrocketing as much as 33% in a single year due to Stargate. And that data center brings in over 6,000 workers just to make sure that building is up and running and providing what it needs to. And as these data centers do go live, do you see that those workers will now move to Abilene full-time, or will this eventually become kind of a self-manned site? Or will they need constant hardware upgrades for the next generations of AI as they continue to come out, and possibly disrupt the towns that they're permanently staying in?

Amy Gromowski: Yeah, it is important to think about what is disruption and what's the long-term effect, if there is any, of that disruption. So for towns like Abilene, while the pitch is that, you know, this tax windfall is going to secure the municipality's financial future for a full generation, the arrival of—you said 6,000, you know, I think on average it's 5,000 to 10,000 highly paid construction workers and specialty contractors—acts as an immediate shock to a local housing market. And this workforce is transient, but it's also perpetual. There's constant, you know, hardware retrofitting, for example, and that eats up starter homes, right? People are going to be there for a while, so starter homes, apartments, long-term hotel stays—that all is consumed by, you know, this new workforce and the injection of the workforce, leaving local service workers like teachers, nurses, retail staff, they're suddenly priced out of their town, you know, their own town where they've been.

So, you know, the friction happens because the timeline of capital deployment is faster than the timeline of civic planning. A tech giant can, you know, just decide to drop 3 billion in a town in a matter of months, but it takes years for a town to zone, permit, and build the housing and the grid infrastructure to support it, and, you know, we can see that in our data at Cotality. So, you know, a 33% rent spike is the shockwave of the initial buildout, but the ongoing operational needs of a data center, it does still require a crew. You know, there's security, there's cooling tech, just the, you know, the retrofitting I mentioned. And so that disruption isn't likely to end when, you know, the data center is constructed. For example, AI hardware, my understanding of it, you know, it has a short life—it's around one to two years. So these facilities will exist in a state of perpetual retrofitting, and that workforce, you know, some portion of it will need to stay, and so that creates a new sustained floor for rental demand and housing demand.

Allie Barefoot: Yeah, I mean this has all been super interesting, Amy. I have a couple more questions for you, but let's bring it back down to a single force: the homeowner, you know? And what happens when, like you said, they've lived in this town forever, this is their home, this is where they're raising their children, and then boom—a massive gray box server farm is now down the street, hypothetically. And they start to think about their property value. However, Cotality data suggests an infrastructure halo, possibly, rather than just a decrease in your property value. So how does a $3 billion power and fiber grid upgrade actually boost a home's automated valuation model, excuse me? And is the cloud actually making land beneath it more valuable?

Amy Gromowski: It sounds counterintuitive because people do fear the gray box, but the data is proving that the halo effect is real. We can see that in Cotality data, but I'll also just, you know, point to a 2025 study that George Mason University did looking at North Virginia—I'm sorry, Northern Virginia—data center alley. And it found that homes, you know, closer to data centers actually sold for more and not less. So why is that? Because you can't build a massive AI factory without upgrading, you know, a lot of the surrounding bedrock. You know, I mentioned earlier the investment in neighborhood schools, parks, police and fire services, utility companies need to invest there, right? So there's a massive upgrade in that area, in that community, you know, the generation of a lot of demand. So just imagine what that's doing for the local economy as the influx of people come in and the influx of dollars, right? Into those municipalities. So we can really see from Cotality's AVMs—our automated valuation models—these systemic, you know, neighborhood improvements do drive up property value.

Allie Barefoot: Wow. You know, it's super interesting once you dive deeper into that actual data. And when we think of data centers being these mega projects, they use the same contractors and materials as large residential buildings. And we're seeing the data centers outbid local builders for electricians and materials. So how can Cotality's geospatial mapping help a home builder see around the corner, essentially, to know if a local grid or labor pool is about to hit its breaking point?

Amy Gromowski: Yeah, builders, municipalities, all of the entities within this ecosystem—they're going to win, we're all going to win when we use predictive data, you know, insights that we build at Cotality to demand smart growth. So our geospatial intelligence really becomes a builder's and a municipality's best defense, right? So what we can do is we can allow home builders to see the collision course before it happens. If our map shows a $2 billion data campus breaking ground—a $2 billion data campus breaking ground in 12 months—then a residential builder knows they need to sequence their builds now or, intentionally to what you were saying in this question, intentionally source their contractors maybe from two counties over and plan for that accordingly, right? To avoid this labor vacuum.

Allie Barefoot: And we have taken a deep dive into some data here, Amy, during this conversation, but let's look a little bit beyond the spreadsheets and the geospatial maps. You know, my last question I have for you is: what is your hope for how this technology can actually help us and homeowners build smarter communities, rather than just bigger data centers?

Amy Gromowski: My hope is that we transition from being reactive to predictive. Right now, we build a data center and we react to a 33% rent spike that it causes. But using the very AI these centers power—interestingly enough, right?—we can build predictive urban models. If we know that a 1.2-gigawatt facility is coming, AI can help city planners precisely map out the necessary housing, transit, and retail footprints required to support the workforce before the affordability crisis hits. So we can make the cloud work for the neighborhood, not just in it.

Allie Barefoot: I completely agree. Amy, thank you so much for breaking down all of this data with us, talking about AI, and I'm sure we'll have another conversation as AI continuously evolves. But thank you again for your time.

Amy Gromowski: Thank you, Allie.

Allie Barefoot: Thank you so much to Amy Gromowski for joining us on Data in Context, and thank you for listening. If you haven't already, subscribe to Cotality's YouTube page, and if you want to find out more information, as always, head on over to Cotality.com