Despite being in ‘top of the first inning’ with AI, a new survey shows just how far consumer trust in AI has come
Brokers and lenders have spent the past year racing to adopt AI tools, often to stay ahead of where they assume their customers still are. New survey data suggests many buyers are already further along than that assumption gives them credit for.
A new LendingTree survey found that 72% would use AI for at least one task somewhere in the homebuying process. The most common uses were searching for homes within budget, estimating a home's value, and finding down payment assistance programs.
The analyst who oversaw the survey said the results caught him off guard.
Matt Schulz (pictured top), chief consumer finance analyst at LendingTree, has spent the past year watching AI work its way into nearly every stage of the mortgage business. He didn’t expect so many to be willing to adopt the new tech so quickly.
"I was surprised to a degree that as many people said that they'd be willing to use it in the process, because we're in the top of the first inning as far as AI goes," Schulz told Mortgage Professional America. "To have three-quarters of people say that they'd be willing to use it to some degree, I was a little surprised by, especially because with any technology, there's always skepticism and there's always a segment that is concerned and just kind of leery of jumping in."
Still improvement needed
Schulz said he was surprised because while AI is advancing quickly, it is still working through the kind of early adoption curve most new technologies go through before they mature.
One of the biggest criticisms of AI-chat programs is the seeming desire to return an answer that satisfies the user, sometimes even at the expense of the truth.
"It's also concerning to me because AI in these early stages is such a people pleaser, and also it can just be flat out wrong," he said. "Eventually it'll get better and smarter over time, but still, just the fact that so many people would rely on AI this early to the degree that some people say they would is eye-opening at a minimum."
That willingness extended even further once the survey asked about handing over the entire process. The survey showed 37% of respondents said they would let AI complete a home purchase from start to finish with minimal human involvement, while 41% said they would not and 23% were not sure.
"That number was really shocking as well, in part just because we're just not there yet," Schulz said.
Where humans matter most
Consumers were far more cautious when the survey asked who they would actually trust to help set an offer price. A real estate agent topped the list at 25%, followed by their own research at 23%, while an AI tool trailed at just 9%.
That gap, Schulz said, shows up most clearly at the exact moments where a wrong call costs real money.
"A home purchase is still very much a human process, at least at key points," he said. "You can do a lot of the run-up, and that sort of thing with AI help, and it can be really useful, but there's a whole lot of steps in it that need that personal touch and need a human being to be involved."
The survey also asked what specifically holds people back from trusting AI more with their home search. The largest share, 22%, pointed to the risk of getting bad or misleading information, followed by data security concerns at 16%.
Looking further out, the survey found consumers split on how much AI will ultimately take over. The survey showed 36% said they expect AI to handle most or nearly all of the homebuying process by 2036, while 64% expect it to play a smaller, supporting role even a decade from now.
Schulz said that same openness to new technology showed up elsewhere in the survey, including in a question about 3D-printed homes. Nearly half of respondents, 49%, said they would consider buying one.
"I think a good amount of that number is just people thinking hypothetically or theoretically down the line," he said. "I think it's just another example of how people are willing to consider an awful lot of things in the name of trying to make a house affordable."
Schulz said the same instinct that makes people open to AI handling parts of a mortgage is also pushing them to consider construction methods that barely existed a decade ago.
"This is definitely less about a fascination with the idea of a 3D-printed home and more about an interest in something new that might be able to make housing a little more affordable," he said.
Stay updated with the freshest mortgage news. Get exclusive interviews, breaking news, and industry events in your inbox, and always be the first to know by subscribing to our FREE daily newsletter.
