Every property person I know was raised on the same three words. Location, location, location. I was too. For most of my career it held up so well that nobody bothered to question it. You could get almost everything else wrong about a building and a great address would forgive you.
But anyone who has spent enough time in real estate knows the address only gets people through the door. What keeps them there is harder to measure. How the lobby feels at eight in the morning. Whether the lift comes when you need it. Whether someone answers when the hot water stops on a Friday evening. Real estate has always been an experience business pretending to be only a location business. We just never had the tools to manage the experience as carefully as we managed the rent roll.
That, I think, is what AI changes.
The occupiers are already telling us. JLL’s 2026 Future of Work survey found that 62% of large occupiers would now prioritise building quality and amenities over prime location, and 66% prefer AI-enabled buildings to ones with basic building management. Obviously this is more relevant for offices and less for other asset classes. Location hasn’t stopped mattering and never will. But for the first time in my working life, it has a serious competitor: the experience of being inside the building.
Most of the AI conversation in our industry is about owners cutting costs. That’s real, and I’ll come to it. But the more interesting story is what AI allows a building to feel like, and how that feeling eventually finds its way back into value.
What changes for those who own the building
On investment, the most useful research I’ve read this year is JLL’s work with MIT, published in July. AI isn’t shrinking real estate demand evenly. It’s splitting markets apart. In San Francisco, a city with some of the highest exposure to AI-driven job losses in the US, nearly 30% of all leasing since 2025 has come from AI companies. There is something interesting in that contradiction. The same technology that is reducing demand in one part of the economy is creating it somewhere else, sometimes in the same city.
For investors, the challenge is seeing those shifts early enough. By the time a trend is obvious in comparable sales, much of the opportunity has usually gone with it.
On operations, Morgan Stanley studied 162 REITs and real estate firms and estimated that 37% of the tasks they perform could be automated, worth around $34 billion in efficiencies by 2030. Most of it isn’t glamorous. Lease administration, reporting, tenant communication, maintenance scheduling. The work that quietly eats a property team’s week.
And then there’s the humbling number. JLL found that 92% of corporate real estate teams are piloting AI or plan to this year, yet only 5% say they’ve achieved most of their goals. I’ve watched enough technology programmes to know that the gap is rarely just about the technology. It’s about whether anyone was willing to change how the team actually works.
What changes for those who live in it
A home is the most emotional asset class there is. People forgive a lot at work. They forgive very little at home.
Think about how renting has felt for most people. You find a listing on a Sunday night, send an enquiry and wait until Tuesday for a reply. You take time off to view a flat with an agent who’s running late. You move in, the boiler fails in January, and you spend two days chasing someone who can authorise a repair. None of that is really about the building. It’s about the feeling of not being looked after. And that feeling is what people remember when their lease comes up for renewal.
Responsiveness is exactly what AI is good at. In the US it’s already happening at scale. EliseAI, one of the leading platforms, says it now handles leasing and resident communication for roughly one in six apartments in the country. An enquiry at 11pm gets an answer at 11pm. Viewings can be self-guided, with the door unlocking from the renter’s phone. Repairs get triaged, scheduled and followed up without anyone having to chase. For the resident, it feels like the building is finally paying attention.
The most revealing research on this came from residents themselves. HappyCo commissioned an independent study asking renters what they actually want from AI. Sixty-seven percent were comfortable using it for general questions. That fell to 53% for routine maintenance, 36% for rent and billing, and just 30% for reporting an emergency. When something urgent goes wrong, 82% want a human first.
I think that’s exactly right. Residents don’t want a chatbot. They want to feel looked after. Let AI do the work behind the scenes, routing the request, finding the contractor, sending the update, so that when a person does pick up the phone, they already know what’s wrong and when it’ll be fixed.
Put the machine at the front door and the human at the back, and you get the worst of both. Do it the other way round and residents get something they’ve rarely had: a landlord who responds faster than they expected.
For Europe, and especially for the Mediterranean markets I know best, this is still early. Much of the rental market here still runs on phone calls, WhatsApp and personal relationships. That’s partly charm and partly inefficiency. The institutional build-to-rent operators arriving in European cities will bring these tools with them. Once residents experience that level of responsiveness, I suspect their expectations of everyone else will change too. Smaller landlords should think about that now, not when their tenants start comparing.
Every building is becoming a host
The same shift is happening across other asset classes, just in different ways. In offices, JLL talks about “hospitality-grade” buildings: places that know how many people are coming in, adjust the air and light to match, and make arriving, booking a room or bringing a guest feel effortless.
Hotels have always understood this better than the rest of real estate. They sell experience as much as they sell rooms. AI makes remembering a guest’s preferences cheap enough to do for everyone, not just the regulars.
Even in logistics, where the occupant is as much a machine as a person, the experience that matters is reliability: the power, uptime and data needed to keep increasingly automated operations running.
The applications are different, but something fundamental is changing. We are starting to expect buildings to respond to the people using them rather than asking people to adapt to the building. And the market is already showing signs of rewarding the best spaces. JLL’s research shows office construction in the US and Europe at historic lows while rents for the best buildings are at record highs. The gap between the best spaces and everything else is widening.
Where the two sides meet
The owner’s AI and the occupant’s experience are really the same thing, seen from two sides. A repair fixed before it becomes a problem is a cost saved on one side of the ledger and a moment of trust on the other. Trust can become renewal. Renewal means fewer empty months and less spent on leasing. Eventually, that finds its way into value.
Two buildings on the same street, with similar specifications, can already perform very differently. Increasingly, part of that difference will come down to how they feel to live and work in. I’ll admit I don’t yet know how quickly valuers will start pricing that in. But I’m fairly sure the market will get there before the valuation frameworks do.
A building that learns
The idea I keep coming back to isn’t really about property management. It’s about what a building is.
For most of history, buildings have been among the most static things we make. Designed once, built once, and then slowly falling behind the people who use them. A building never knew whether it was full or empty, too warm or too cold, loved or merely tolerated. It just stood there, and we adapted to it.
What AI introduces, for the first time, is the possibility of a building that learns. One that notices patterns, remembers preferences, anticipates problems and gets better the longer people use it.
Think about a hotel you’ve stayed in more than once. The second visit feels better than the first because they remember you. The room is ready the way you like it. Nobody asks the same questions twice. Until now, buildings couldn’t really do that. Every tenant, every resident, every visitor started from zero, every single day. A building that learns remembers. The fabric still ages, but the way it serves the people inside can get better every year.
I’m not sure yet what that means for how we design, finance or value property. My instinct says it will change all three. And the people who figure it out first won’t necessarily be the ones holding the best addresses.
Location will always matter. But the next decade of real estate value will be decided by something the old mantra never mentioned: how intelligently a building treats the people inside it.
Location, location, intelligence.