From a robot that can handwrite hundreds of real estate prospecting cards in minutes to a digital consultant that can narrow down thousands of mortgage options for a client, artificial intelligence has finally arrived in the real estate industry.
Across Southern California, housing experts, mortgage lenders and real estate agents have taken notice of the technology’s ability to speed up transactions in a process once thought to be time-consuming and stressful.
JJ Mazzo, a mortgage lender with CrossCountry Mortgage in San Juan Capistrano, said that he can pinpoint precise mortgage options for certain clients’ needs in a matter of minutes with an AI program, Mortgage Coach.
For instance, the AI software produces side-by-side comparisons on different loan variables, highlighting interest payments, tax benefits and cost estimates. It also breaks them down for clients using interactive graphs, charts, and video narration to explain complex mortgage procedures clearly.
Before using an automated system, his company produced a larger set of options by hand.
“I can find somebody’s needs and wants and be able to put them into a format that’s much easier to present to even a prospective loan officer I want to work for me,” said Mazzo of the new AI program.
For many homebuyers, real estate comes with financial risks and complexities, and being able to use an AI-generated presentation that translates the details clearly is helpful to Mazzo and his clients.
Real estate agents also are trying different ways to expand their reach using AI.
Paul Young, an Irvine-based agent, said that a new AI-powered robot called Handwrytten handwrites hundreds of prospecting cards that he sends to potential sellers.
While he previously wrote out the cards himself, he wasn’t able to produce nearly the same number at the rate of the machine, and has seen the added production rate triple his client responses. That, in turn, improved his grip on “geographic farming,” the strategy agents use to carve out a presence in targeted neighborhoods.
“I can only write 10 cards a day,” Young said. “This service has no limit. You can scale as much as possible.”
AI, which has recently been adopted by listing sites Zillow and Redfin, could help consumers find more specific properties, said Davide Proserpio, an assistant professor of business administration at USC.
“AI is going to help increase the number of searches and also increase the number of tours that the consumer asks for. Searches are going to be easier to perform,” Proserpio said.
He added that AI companies tend to use the latest models for new products, and large language models are most common right now.
Because large language models contain vast datasets that can contain biased information, there are risks in basing real estate decisions too closely on what technologies suggest.
He said research shows that minorities could be discriminated against and simply denied in a mortgage application because the data on which these AI tools are trained is biased against them.
“We know from the past that the application approval process can be biased even when driven by a human,” Proserpio said. “Now we put an algorithm in the middle and things could get even worse.”
Since October, the California Civil Rights Department has criminalized housing discrimination based on AI tools.
The department found that an AI automated-decision system could violate state law if it harms applicants or employees based on protected characteristics, such as gender, race, or disability.
Proserpio says the degree to which these technologies could unlearn the biases at all remains an open question.
Others say AI tools could add interpersonal disagreements that hold up real estate contracts.
Los Angeles real estate agent and coach Paul Figueiredo thinks the real estate industry is an inherently social one, and no matter how deeply AI technology is embedded into the client-partner relationship, people will be necessary to mitigate disagreements.
He points to the tendency of property buyers, sellers and agents consulting price points with ChatGPT, and how respective responses can contradict one another and end transactions.
“Somebody who has just a little bit of information can turn [a deal] sideways for them and ruin their confidence,” Figueiredo said. “Now people have that at their fingertips.”