AI in Real Estate: JLL’s Strategy and Reality Breakdown

HOST

From DailyListen, I'm Alex. Today: how JLL is navigating the massive shift toward artificial intelligence in real estate. It's a field that’s seeing huge investments, but there's a gap between the hype and the reality. To help us understand, we have Marcus, our economics analyst, who's been covering this for us.

MARCUS

Thanks for having me, Alex. It's a pleasure to be here. You’re right to highlight that gap. When we look at JLL’s recent data, we see a company—and an entire industry—trying to reconcile a 243-year legacy with the breakneck speed of AI. The core of this story is the $3 trillion supercycle in data center infrastructure. By 2030, we expect to see 100 gigawatts of new supply. That’s a staggering amount of power and physical capacity. AI is the primary catalyst here. It’s projected that AI will account for half of all data center workloads by the end of the decade. But as you noted, there's a reality check. While 90% of companies are running AI pilots, only 5% have actually hit their goals. It’s a classic case of the technology moving faster than the internal infrastructure, data quality, and change-management processes needed to make it work.

HOST

Wow, that’s a pretty wide gap between 90% piloting and only 5% succeeding. It sounds like a lot of companies are just throwing technology at the wall to see what sticks. Does this mean these companies are failing, or are they just in the messy, early stages of a long transition?

MARCUS

It’s definitely more of a messy transition than a total failure. Most organizations, including those in corporate real estate, are running about five AI use cases at once. They're trying to standardize data, detect anomalies in building operations, and automate reporting. The issue isn't necessarily the technology itself, but the organizational readiness. If your underlying data is fragmented, AI won't magically fix it. It'll just process bad data faster. Successful companies are moving beyond the "pilot everything" phase and focusing on specific, value-driven outcomes. They treat their buildings not as static assets, but as dynamic platforms that interact with power markets and public policy. Those who get it right are the ones who can handle the complexity of integrating AI into their core workflows. It’s about building the foundation before you scale, and that’s where most of the current friction lies.

HOST

So, the foundation is the problem, not the AI itself. That makes sense. But let's talk about the data centers, because that’s where the real money is moving. You mentioned a $3 trillion supercycle. That number is huge. Why is the cost to build these things rising, and how does AI influence that?

MARCUS

The cost is rising because we aren't just building warehouses; we're building power plants. Data center construction costs have been climbing at a 7% compound annual growth rate. A big part of that is the requirement for AI-optimized facilities. These aren't your standard server farms. They need extreme power density and advanced cooling systems that can handle the heat generated by high-performance AI chips. Because these facilities are so specialized, they can command lease rates up to 60% higher than traditional space. It’s a total shift in the market. Operators who can secure power and deploy these specialized structures quickly are the ones winning. And that’s where the power delivery comes in. It used to be a constraint, but now, having a reliable, standardized power strategy is a massive competitive advantage. You can't run the AI revolution on a grid that wasn't designed for this kind of load.

HOST

That’s a clear explanation of why those rents are so high. But I have to ask: what’s the risk here? If everyone is betting $3 trillion on this, what happens if the AI demand doesn't actually materialize at that scale? Is there any criticism or concern about this massive build-out?

MARCUS

That’s a fair question. There’s definitely a risk of overbuilding if the demand curves shift. While AI is driving massive investment, we’re also seeing energy constraints that could slow down deployment. Some critics argue that the industry is ignoring the environmental and regulatory hurdles of such rapid expansion. If the power isn't there, or if local communities push back on the massive energy demands of these centers, that $3 trillion might not yield the expected returns. Furthermore, there’s the issue of talent. You need specialized engineers to build and manage these high-density facilities, and there’s a shortage globally. If you can’t staff the buildings, the physical infrastructure becomes a liability rather than an asset. It’s a high-stakes game. Investors are currently prioritizing speed to market, but that strategy assumes the tech market will remain as hungry for capacity as it is today.

HOST

That risk of overbuilding is a point I hadn't fully considered, especially with the talent shortage. So, if we look ahead to 2030, with this 14% compound annual growth rate, where is this growth actually happening? Is it global, or are we just talking about a few major tech hubs in the US?

MARCUS

It’s truly a global phenomenon, though the intensity varies. While the US is a massive market, we’re seeing significant growth in Europe and key parts of Asia. The demand is driven by where the compute is needed and where the power can be accessed. Think of it as a search for the path of least resistance. You need proximity to the users, but you also need a grid that can handle the load. In some regions, we see operators turning to on-site power and battery storage just to get projects off the ground. It’s not just about finding a plot of land anymore. It’s about securing the energy ecosystem. This is why the next phase of growth will favor those who can navigate the intersection of public policy, power markets, and technical requirements. It’s a complex landscape, and geography is just one piece of the puzzle.

HOST

That makes sense. It sounds like geography is secondary to power availability. Now, you mentioned that 92% of occupiers are running AI pilots. For those listening who might be in corporate real estate, what are they actually doing with it? Are they just automating reports, or is it deeper than that?

MARCUS

It’s a mix, but it’s definitely moving toward deeper operational insights. Many occupiers are using AI to standardize data across their portfolios. If you have 500 buildings in 50 cities, just getting a consistent view of your energy usage or maintenance costs is a massive challenge. AI helps them aggregate and clean that data. Once you have that, you can use it to detect anomalies—like an HVAC system that’s running inefficiently in the middle of the night. It’s about moving from reactive maintenance to something more proactive. They’re also using it for portfolio analysis, trying to figure out which assets to keep, renovate, or divest based on future performance projections. It’s not just about saving money on admin tasks; it’s about making better, data-backed decisions on how to manage physical space in a world that’s changing quickly.

HOST

So, it's about efficiency and better decision-making. But with all these pilots happening, have we seen any actual, measurable value yet, or is it still just "potential"? I want to know if the 5% who succeeded are actually seeing a return on their investment, or if it's just hype.

MARCUS

The ones who are succeeding are definitely seeing a return, but it’s usually in operational efficiency. We’re talking about reduced energy consumption, lower maintenance costs, and faster reporting cycles. The issue is that the value is often incremental, not a sudden, massive windfall. It’s hard to point to a single "AI success" and say it paid for the entire investment in one quarter. Instead, it’s a series of small, significant gains. The companies that are getting it right are the ones that don't view AI as a magic button. They view it as a tool to gain a more holistic understanding of their operations. They’re preparing for waves of change we can't fully predict yet. It’s a long game. The 5% who hit their goals are those who invested in the underlying data quality first, rather than just buying the latest software.

HOST

That’s a helpful distinction. It sounds like the "boring" stuff—data hygiene and organization—is actually the most important part of the AI strategy. Before we go, what should someone watching this space look for over the next year? Are there any signals that things are stabilizing or changing?

MARCUS

Keep an eye on how these companies handle the power constraint issue. That’s the real bottleneck. If we see more partnerships between data center developers and energy providers, that’s a sign of maturity. Also, watch for consolidation. With construction costs rising and the need for specialized infrastructure becoming so intense, smaller operators might struggle. We could see a shift where only the largest, best-capitalized players can afford to play in the AI-optimized space. Finally, look at the change-management side. The companies that successfully integrate AI are the ones that retrain their people to work alongside these systems. If you see a company investing as much in their workforce as they are in their servers, that’s a very strong signal that they’re in it for the long term.

HOST

That was Marcus, our economics analyst. The big takeaway here is that while AI is driving a massive $3 trillion supercycle in infrastructure, success isn't about the tech itself—it's about the organizational foundation, power availability, and long-term planning. It’s a messy, expensive transition, but one that’s reshaping the future of real estate. I'm Alex. Thanks for listening to DailyListen.