AI has entered the real estate chat—and it doesn’t appear to be going anywhere anytime soon.
While some say the AI bubble may burst in the next year or two, that’s not the feeling you get perusing real estate tradeshow exhibit halls where vendors pitch AI solutions for just about everything.
Whether it’s the promise of streamlining agents’ marketing, lead generation, writing listing descriptions and photo captions or tackling other less-than-enjoyable administrative tasks, emerging AI products can seemingly do it all.
The goal? Help busy real estate agents reclaim their most precious commodity: time.
But how exactly is the promise of AI for real estate stacking up? And where is it falling short of expectations?
“I absolutely 100% believe that brokerages and agents and really anyone running a business needs to be all-in on AI or they will not be competitive in two or three years,” says Malte Kramer, founder and CEO of Luxury Presence. Still, agents and brokerages should take care to vet products and vendors to “make sure you’re working with the best-in-class products.”
This tension between promise and peril defines real estate’s current AI moment.
Although 68% of real estate agents say they’ve adopted AI in their business, only 17% said it’s had a significant positive impact, suggesting a gap between adoption and results, according to the National Association of Realtors®’ (NAR) 2025 Realtor Technology Survey.
One thing’s for sure: everyone is using AI (even consumers). A Realtor.com® survey found that 82% of consumers are using AI to gain intel about the real estate market. Consumers said they mostly used ChatGPT (67%) and Gemini (54%) in their AI-guided home research.
Brokerages are going all-in on the technology, too. In December 2024, SERHANT raised $45 million in equity funding to further develop its AI-powered S.MPLE platform. The Real Brokerage, which closed nearly $50 billion in sales volume in 2024, attributes much of its success to its forward-looking investments in AI platforms such as Real Wallet (a fintech platform) and other AI-powered tools to support agent productivity—even going so far as to swap its CEO with an AI stand-in on an earnings call.
Countless others are beefing up investment and making strategic hires to steer their real estate brokerages into the future—a future where AI will touch virtually everything we do.
Where AI gets it right
When vetted and deployed correctly, AI can deliver strong results that are worth the investment—and live up to its grand promises.
For instance, Kramer’s company’s AI marketing specialist has written and published over 50,000 blog posts, qualified over 70,000 leads and made over 15,000 SEO optimizations in just 60 days. One brokerage client reduced marketing expenses by 80% while simultaneously increasing website traffic sixfold and engaged leads by 50%, Kramer claims.
The key to these impressive results? What Kramer and others call a “human in the loop.” This means that AI handles the bulk of work while marketing experts oversee quality control and compliance.
“The approval rate for those 50,000 blog posts from our customers, from real estate agents, is over 95%,” he says, emphasizing that properly built systems can meet professional standards.
The types of tasks that are well suited to AI (or any kind of automation) are tasks that are repetitive, dangerous or dull. In real estate practice, this includes administrative tasks, says Nikki Greenberg, a real estate technology strategist and keynote speaker.
Elise AI, which features AI call centers and chatbots for property management, has been a game-changer in the real estate industry, she says.
Experts gave rave reviews to other AI products for real estate: lead-gen platform Scout and Purlin, a suite of comprehensive marketing solutions powered by AI and built for real estate. The suite is also compliant with Fair Housing rules.
As the technology gets better, it will take on more complex tasks, but we’re not quite there yet, she cautions.
“The algorithms are only as good as the training and, unfortunately, it tends to repeat historical biases,” Greenberg says. “You have to have a person supervising what’s coming out.”
For top-producing practitioners like Ravi Kantha, a team leader with SERHANT in New York City, AI’s value lies in the vast amounts of data it can synthesize and translate into actionable insights. “One of the problems in our business versus a lot of industries is that the data is scattered,” Kantha explains. “There’s so much of it that’s publicly available, but it’s in 150 different places.”
Kantha says his team invested about $125,000 this year building a proprietary database that consolidates information on properties, potential affluent buyers and market trends. Using AI to organize this data has resulted in more than 90% accuracy in identifying buyer pools.
“That’s something that was not even remotely possible five years ago,” Kantha says.
Kantha’s team can now quickly hyper-target their marketing with more precision. And when the team put their own spin and voice into rewriting generic AI-generated text campaigns, response rates jumped from single digits up to 37%, yielding several qualified meetings with prospective luxury home sellers.
Where AI overpromises but underdelivers
For every success story, the industry is littered with disappointments. The problem, multiple leaders agree, comes from vendors dressing up out-of-the-box products with “AI-powered” labels.
“The majority of AI tools I’ve seen get launched are sort of AI assistants,” Kramer says. “They sit on top of OpenAI and they have some basic prompts built in, and they don’t really leverage a lot of underlying real estate data effectively. What you end up with is a pretty superficial use case.”
Ashley Stinton, managing partner of NAR REACH, the trade group’s technology accelerator, sees the same pattern. “I think AI as a buzzword has become a little bit overhyped, and that can be noisy,” she says. “Everything is AI now. To me, that’s where I would like to see the hype kind of calm down a bit.”
Lead generation and CRM tools have been ground zero for this chatbot “wrapper” phenomenon. Tools that merely add AI-generated copy to existing platforms fail to deliver transformative results because they produce generic outputs that don’t reflect individual agent brands or market nuances.
To get more targeted results, you need a custom-built AI technology, says Michael Thorne, a sales representative with REMAX Lifestyles Realty in Langley, British Columbia, Canada, who trains real estate agents on using AI.
“You are going to pay a company $10, $15 or $20 a month for a vanilla prompt,” Thorne says. “You are going to get the average.”
As AI takes over, an authenticity crisis emerges
The average output creates a clear authenticity problem for an industry that’s built entirely on personal relationships. Real estate consumers—especially those in sophisticated markets—are becoming more adept at spotting AI-generated content, and they don’t love it.
Kantha’s experience confirms this concern.
“When we ran a text campaign with the spammy nonsense that these companies provide to you, we had like an 8% or 9% response rate, and most of it was like, leave me alone,” he recalls.
The solution requires customization that most off-the-shelf tools don’t provide. “If you give the AI a generic prompt, you’re going to get a generic answer,” Kramer explains. “Training an AI on each specific customer is how we ensure that when they’re nurturing a lead, they sound like that brand.”
Successful AI implementations must be done transparently—or you risk alienating your customers, Thorne points out.
“If the public believes it’s you, and it’s not you, I think that is career ending,” Thorne warns. “If you are baiting and switching, if you are tricking your client into believing they’re having a conversation with you, and they later find out it’s not you, that’s different.”
Where human connection can’t be replicated in real estate
One thing everyone can agree on: AI should never replace core relationship-building activities.
“My wife’s a real estate agent, and when I see what she does every day—she hosted a big community event here in the neighborhood, and you look at the relationships that are being built, the level of trust—there is no world where that gets replaced by AI,” says Kramer.
Thorne contends that AI should handle what he calls “$20 an hour work” so agents can focus on “$300 an hour work,” or the high-touch, high-value client interactions that lead to repeat and referral business.
According to NAR, Realtors® perform 179 tasks over the duration of a transaction—from initial client outreach through closing. According to WAV Group, a real estate consulting firm, ChatGPT asserts it can complete 110 (or 60%) of these tasks. But the remaining 40% are and will continue to be agent-driven.
Whether it’s a young first-time buyer of a starter home or a sophisticated high-net-worth seller of a multimillion-dollar mansion, AI can’t replicate the reassurance and facetime agents provide their clients one on one.
Vetting AI real estate tools—with compliance in mind
More agents and brokerages are integrating AI solutions into their tech stacks. But before they do, they have an unenviable and cumbersome task: vetting the products.
Kramer recommends a four-step framework. First, evaluate vendor talent and engineering capabilities.
“If it’s two guys in a garage, it’s going to be difficult (to support you),” he notes. Second, ask for simple explanations of how products work beyond ChatGPT prompts. Third, review data privacy policies and compliance measures. Finally, request case studies and conduct real-world pilots, Kramer recommends.
Don’t underestimate the power of peer validation, too, Stinton says. “In this industry, peer to peer very much matters,” she says. “Not just the case study of ‘it works and I really like it,’ but quantify what that means back to your business.”
NAR REACH, for instance, focuses on companies like Scout, which uses predictive analytics to match agents with high-propensity sellers, and Perlin, which has developed a Fair Housing-compliant suite of AI solutions.
“They’re training on data that is specific to our industry,” Stinton explains, addressing a critical compliance concern.
Fair Housing violations and hallucinations—AI’s tendency to confidently state (or make up) incorrect information—pose significant liability risks for agents and brokerages.
“Don’t outsource your thinking to AI,” Kramer advises. “Treat the AI like an employee. It can do the work for you, but you have to check the work. You’re ultimately responsible for the output.”
This responsibility extends to emerging applications like AI photo staging, which can create misleading property representations.
“You’re making a representation of an actual property that AI is guessing at,” Thorne says. “There’s a lot of ethical questions we have to be asking ourselves.”
Going all-in vs. a measured approach
AI investment depends on your needs, annual sales volume and business model. While Kantha’s team makes a significant annual investment with demonstrable ROI, most agents should start far smaller, Greenberg says. Look for quick wins using tools like ChatGPT to save you time while you pursue longer-term transformational data projects, she adds.
Thorne agrees with this incremental approach.
“AI is very good at getting you from zero to 80% very quickly,” Thorne says. “But the next 15% is your responsibility. Is it factual? Does it speak to my brand? Is it correct?”
Before buying into the AI hype and going all-in on a buying spree, be clear on what problems you need to solve, experts say. The key lies in knowing your time’s value.
“If you can make yourself more efficient, that’s worthwhile in the first place for any agent,” Kantha notes. Calculate how much an hour of your time costs your business, then assess whether AI tools save enough hours to justify their expense, he adds.
What’s next for AI in real estate?
Looking ahead, industry leaders envision a real estate reality where AI enables agents to serve more clients more effectively. But it won’t ever replace them.
“It will make the businesses that deploy correctly massively more efficient,” Kramer predicts. “It will give real estate agents time back to spend with their families and not be as stressed out at night or on the weekends.”
Thorne sees an even more dramatic shift coming with “AI agents.” These are autonomous systems that work independently once trained.
“You’re going to have 30 or 40 of these virtual employees that you’ve trained working in your business, doing very specific tasks, doing it perfectly every single time,” says Thorne, adding that it will free up your time to focus on relationships and sales.
The transformation won’t happen overnight. Real estate is known to move at a glacial pace when it comes to embracing new technology—and it’s still early days. Think of it as “a teammate, not a tool,” which requires ongoing training and oversight, Thorne says.
But make no mistake: AI in the real estate industry is here to stay. Finding a balance in how (and when) you use it is the challenge ahead.
“I don’t know how an agent that ignores this competes,” Thorne says. “But at the same time, agents that put this technology between them and their client will lose.”
