Michael Reisor can usually tell within a few sentences that he’s not reading a homebuyer’s own words. The verbiage, formatting and diction give it away.
“We all know that’s ChatGPT,” said Reisor, a Compass agent who runs the Reisor.Team across Texas and New York. “We just know what’s going on. But I think a lot of older people feel like, ‘I’ve got the upper hand; there’s no competition with me. I’ve kind of got the world on my side.'”
What Reisor is describing isn’t buyers Googling before a showing, but rather uploading 60- to 80-page inspection reports directly into ChatGPT or Claude, then forwarding the AI chatbot’s summary back to him as a negotiating position.
He estimates it’s happening in 10 percent to 20 percent of his deals right now, and it’s most pronounced in a specific scenario: parents funding a purchase for adult kids.
The Great Wealth Transfer meets the chatbot era
With the transfer of wealth from baby boomers to millennials and Gen Z, Reisor said his team is seeing more parents involved in transactions than ever before. Various studies and surveys bear this out.
Nearly three-quarters of parents with kids still at home say they’re either already saving or planning to help their children buy a home someday, according to Northwestern Mutual’s 2026 Planning & Progress Study.
Nearly one in five first-time buyers say they plan to lean on their parents for financial help to purchase a home, according to a 2024 survey by Intuit Credit Karma. That share nearly triples among Gen Z respondents, 44 percent of whom expect parental assistance, compared with 16 percent of millennials and 12 percent of Gen X.
The pattern holds among those who’ve already bought: 19 percent of current or former homeowners say they received financial help from their parents, including 38 percent of Gen Z buyers and 27 percent of millennials.
The fact that some of these parents are turning to AI for housing guidance isn’t much of a surprise. A Realtor.com survey last year put a number on how far AI has crept into house hunting: 82 percent of Americans say they’ve used it for real estate insights.
When respondents were asked what makes them feel “smarter” about the market, real estate agents topped the list at 62 percent. But AI wasn’t far behind at 61 percent, a near-tie that may help explain the friction agents like Reisor are running into.
Baby boomers trust the chatbot. Their kids don’t
Reisor first noticed the pattern in deals where a parent was supplying gift funds or co-purchasing outright. In his experience, younger buyers tend to treat chatbot output as one input among several, to be weighed against what their agent is telling them.
Older buyers, he said, are more likely to treat it as settled fact and act on it directly, including copy-pasting AI-generated language straight into a negotiation.
“A lot of younger people wouldn’t copy-paste directly what’s coming from chat and send it to their broker as ‘This is what we need to say,'” Reisor told Inman. “But the generalized baby boomer will do that. They’ll just copy-paste it and be like, ‘Here’s what we need to ask for, or we’re not moving forward in the deal.'”
The dynamic gets harder to manage because of who’s writing the checks. Adult children who might otherwise push back on a parent’s AI-sourced demands often won’t, Reisor said, because they don’t want to jeopardize the funding.
“The kids aren’t going to say anything to really challenge the parents, because they don’t want to lose the money, or hurt the purse strings,” he said.
That leaves Reisor threading a needle, correcting the record without alienating the client who’s actually paying him.
Diffusing the situation
The friction point Reisor described most specifically is cost estimation. Chatbots summarizing inspection reports don’t have access to what a specific plumber or contractor in a specific market actually charges, so they default to wide ranges.
Reisor said those ranges routinely inflate minor line items into deal-threatening amounts.
“The chatbot will say something like replacing a pipe underneath a sink could cost between $500 and $5,000,” he said, “because it just wants to give broad strokes. It can’t say exactly what it’s going to cost.”
He’s seen the effect scale up on bigger-ticket items, too. It’s the chatbot’s translation of routine findings into worst-case dollar figures that creates obsessive, hard-to-de-escalate concern from clients.
His fix isn’t to dismiss the AI research outright. He said that approach only makes buyers more defensive. Instead, he reframes it as one data point and substitutes his own track record.
“We don’t want to be at odds with our client, but we have to set expectations,” Reisor said. “That’s what I respond with in these transactions — I call it out. I say, ‘I know that you may be researching on your own and finding certain info points, and I think some of them are relevant, and these are things we can dig into, but in my experience, this is going to cost this.'”
Reisor will also connect clients with trades he has worked with on numerous deals and on his own properties, so they can give quotes and explain pricing.
“That’s how I diffuse these situations,” Reisor said. “Of course, you’re not going to get anywhere being like, ‘Oh, this is so dumb. This is chat; they don’t know what they’re talking about,’ because then they just get more adversarial.”
Contracts are another flashpoint
Reisor said the same pattern shows up when buyers feed contracts, specifically Texas’s promulgated forms, into a chatbot and ask what to change. The output, he said, often recommends redlines or legal language that neither he nor the buyer is authorized to insert.
“We can’t be redlining contracts. It’s not in our purview,” Reisor said. “We can’t practice law in any way.”
Reisor said he has to clarify: This is the form used in pretty much every transaction in the state. He can’t redact clauses, and he can insert things in special provisions only if an attorney has drafted that language.
Not every AI use case is a problem, he added. Buyers asking a chatbot for generic pre-listing advice tend to get broadly sound guidance. It’s the specificity gap on cost, contract language and market-specific precedent where Reisor said AI consistently oversteps what it actually knows.
More AI in real estate doesn’t scare Reisor
Asked whether this trend in buyer behavior will fade as AI models improve, Reisor thinks the opposite will happen. He expects the reliance on AI to keep spreading, not disappear, as more buyers and sellers default to a chatbot as a first stop.
He doesn’t see that as inherently bad for real estate agents. He sees it as raising the bar on what agents need to bring to a transaction that a chatbot can’t replicate.
“People can engage with chatbots all they want and get all this information,” he said. “But at the end of the day, they know the specificity and market knowledge is going to come from their agent, not the chatbot.”
He also pushed back on the idea that this is fundamentally new. Buyers assembling their own research before AI just did it through Google, he said, but more slowly and in a more scattered way. What’s changed is the speed and the false confidence that comes with a single, authoritative-sounding answer.
Despite various stories about buyers and sellers completing transactions without an agent and relying solely on AI, Reisor’s experience leads him to believe that an agent’s guidance and knowledge are still sorely needed.
“Chatbots aren’t able to provide the specificity or context-specific answer for most real estate transactions. There’s just no way,” he said. “Every property, every seller personality, every state, city, precinct is different.”
“There are so many personalities involved, especially with multi-generational things,” Reisor said. “And as we all know from many other stories, chatbots are often just flat-out incorrect.”
