Commercial real estate has never been quick to embrace new technology. But as generative AI advances rapidly, waiting on the sidelines is becoming less of an option. The problem is that while many in the industry recognize the need to engage, far fewer know what that actually looks like in practice.
"What I'm seeing right now, from talking to my students, talking to people in the business, is that people are struggling," Lynn McKee, director of the master of science in commercial real estate program at Georgia State's Robinson College of Business, tells GlobeSt.com.
"They know they've got to use this, but they don't know how."
Part of that confusion stems from a basic misunderstanding of how generative AI works. Unlike traditional software, it doesn't follow fixed rules or produce consistent, predictable outputs.
"At the core, it's not a deterministic model," McKee says.
Instead, these systems rely on probabilities, generating responses based on patterns in data rather than a true understanding of accuracy or meaning. That creates a level of uncertainty that doesn't sit comfortably in an industry built on risk management.
"They hallucinate, they make things up, they get things wrong," he says. "If you take the human out of the loop and let these things run on their own, and that's where agents are going, and there is no human in sight, that's a big problem.
"The height of AI and what it can do is so far from the applications on the ground," McKee adds. "There is concern that revenue will not come to cost-justify it. We're not an early adopter business because we take enormous risk on what we do, investing. When it comes to new technologies, eh. Because we've taken so much risk in investing."
Even at the largest brokerages, where resources and infrastructure should make adoption easier, the rollout has been uneven. Firms are introducing company-wide platforms with built-in controls, but that hasn't necessarily translated into meaningful use.
"What should be the 'most successful implementations' is in the largest brokerages," McKee says.
"They're trying to institute on an enterprise level a platform with guardrails. But if you talk to people in those organizations, they say, 'What am I supposed to do with this?' A lot of them are saying, 'I have it, here is this thing,' and they all have cutesy names, but the average person doesn't know where to dive in and use it."
In some cases, the pressure is increasing faster than the understanding. Employees are being told AI will factor into performance, even as they are still figuring out how to incorporate it into their day-to-day work.
"It will revolutionize the business in ways we know and ways we don't, but we have a long way to go from hype to actual application," McKee says. "Just like all technology. It doesn't work out of the box, and we are still in the early rounds."
For now, the more practical path forward may involve less with the tools themselves and more with the groundwork behind them. That means cleaning up data, tightening processes and addressing compliance and security concerns. It may also mean bringing in outside expertise from groups that have already navigated early implementations.
"After you sift through all this stuff, I think it will be good," McKee says. "It will be a tremendous benefit, but it's a long way to get there. I just don't know if the financial markets and hype have the patience to wait for the point until it makes sense."
© Arc, All Rights Reserved. Request academic re-use from www.copyright.com. All other uses, submit a request to TMSalesOperations@arc-network.com. For more information visit Asset & Logo Licensing.