I Used AI to Buy a House and I Still Hired a REALTOR®.

Bill and I just bought a house, and yes, I used AI to research the property before we made our offer. Yes, me. The person who spends a considerable amount of time questioning whether AI is actually delivering on everything we're being promised. I went straight to ChatGPT, and I was genuinely impressed with how much information I could gather before making one of the largest purchases of my life.

I started with the property's history. I wanted to know what it had previously sold for, how the asking price had changed, how long it had been on the market, and what I could learn about the surrounding neighborhood. I explored historical pricing and other publicly available information about the property because I was curious and wanted to learn as much as I could about the home we were considering.

I knew our REALTOR® (shoutout to Jen Harris) would prepare a CMA and provide professional guidance on pricing and market conditions. I wasn't trying to replicate her work or fact-check her recommendations. I simply wanted to access information in a way that was convenient and comfortable for me. I could ask questions as they occurred to me, explore whatever caught my attention, and get answers without having to call or text her every time I thought of something new.

That's what I think is so important about this experience. AI wasn't replacing a service I expected from my REALTOR®. It was giving me another way to access and understand information throughout the process.

The information I received was useful, and I didn't encounter inaccuracies in what I researched. Jen’s work confirmed what I had learned as we worked through the purchase, but that wasn't why I had used AI in the first place. I used it because it was easy, accessible, and allowed me to explore the property at my own pace.

Despite having access to all of that information, I absolutely needed my REALTOR®, and she exceeded my expectations.

Jen coordinated our showing, walked through the property with us, arranged our inspections, helped us evaluate the findings, and made recommendations about what we should be asking for. She prepared comparable market information, understood what was happening in the local market, and negotiated on our behalf based on the conversations we had about what mattered to us. She advocated for us throughout the transaction, and I felt confident that she was representing our best interests.

I couldn't tell you what technology Jen used to prepare her CMA or develop her recommendations. Considering the industry I work in, there's a pretty good chance I'm familiar with most of those tools. What I found interesting was that, as a buyer, I really didn't care. I wasn't evaluating her tech stack. I cared that she understood the market, could explain her recommendations, and was helping us make good decisions.

Zillow released its 2026 Consumer Housing Trends Report this week, and I think the findings are worth paying attention to, particularly in light of my experience.

According to Zillow, 82% of homebuyers use AI in their everyday lives, and 61% of those AI users have also used it during the homebuying process. Among buyers who used AI for home shopping, 74% used it to learn about the homebuying process, 71% used it for market insights, and 70% used it to find homes. More than half, 54%, even used AI to help recommend an offer.

Think about that for a minute. We're not just talking about consumers asking ChatGPT to write a listing description or explain what earnest money means. They're using AI to research markets, evaluate properties, and help inform financial decisions.

I'm a forty-something homebuyer who happens to work in real estate technology, and despite being an AI skeptic, I did exactly what Zillow's research says consumers are doing. I used AI to help evaluate a property before making an offer. I suspect there are plenty of buyers doing the same thing who don't have the benefit of working in our industry.

Here's another statistic I found interesting. Despite all of this AI adoption, Zillow reports that 85% of buyers still used a real estate agent. Among buyers who worked with an agent, 61% ranked help deciding the details of their offer among their three most valuable services, while 49% included negotiating on their behalf.

That certainly reflects my experience. AI helped me understand the property, but Jen helped me understand what to do with that information. She brought market knowledge, experience, and professional judgment to the transaction, along with something AI couldn't provide: someone who was actually accountable to us and advocating for our interests.

There's another finding in the Zillow report that I think deserves attention. Among buyers who hired an agent, only 24% knew their agent was using AI, another 24% believed their agent was using it, and 52% didn't see any evidence of AI use at all.

I find that fascinating because I work in real estate technology, and I couldn't tell you whether Jen used AI or which technology platforms she relied on throughout our transaction. It didn't matter to me. What mattered was the quality of the service she delivered.

We've spent a tremendous amount of time in this industry discussing whether AI is going to replace real estate agents. After going through this experience myself, I think we're asking the wrong question. Consumers already have access to more information than ever before, and AI is making that information easier to find and understand. That doesn't eliminate the need for professional expertise, but I do think it changes what consumers expect from the professionals they hire.

My REALTOR® didn't need to be the gatekeeper of information to demonstrate her value. Her value was in knowing how to interpret that information, apply her experience, and help us navigate the transaction.

I think there's an equally important conversation here for MLSs.

We've spent years talking about MLS data as the source of truth for real estate information. Now we have consumers increasingly turning to AI applications to research properties, understand market conditions, and even evaluate what they should offer on a home. Zillow's report also found that 70% of buyers would trust AI-generated homebuying insights.

That should get our attention.

My experience with AI was accurate, at least based on what I was able to validate with my REALTOR®. That doesn't mean every consumer is receiving accurate information. We know these models can produce incorrect or outdated answers, particularly when they're relying on information gathered from across the internet. If consumers are going to use AI to make decisions about real estate, shouldn't MLSs be thinking seriously about how to ensure those applications have access to authoritative data?

I almost hate to bring up MCP servers because, frankly, I think we've heard enough about them lately. However, this is one of the use cases where I can see real value. An MCP server could provide a way for AI applications to access authorized MLS information through a governed connection rather than relying exclusively on publicly available sources that may be incomplete or outdated.

Of course, there's more to it than simply standing up an MCP server. MLSs have to consider licensing, permissions, security, attribution, and how that information can be used. Not every piece of MLS data should be available to every AI application, and having access to authoritative data doesn't automatically guarantee an accurate response.

Still, I think the opportunity is significant. If consumers are increasingly using AI as part of their homebuying journey, MLSs should absolutely be part of the conversation about where that information comes from and how it gets delivered.

What I took away from buying this house is that AI and my REALTOR® served two different purposes, and both were valuable. AI helped me research the property and become a more informed buyer. Jen helped me evaluate what I had learned, navigate the transaction, negotiate effectively, and ultimately purchase a home that Bill and I are incredibly excited about.

I think our industry needs to spend less time debating whether consumers should be using AI and more time understanding how they're already using it. For agents, that means recognizing that the value you provide extends well beyond access to information. For MLSs, it means thinking about how your data remains authoritative and accessible as consumer behavior evolves.

Consumers aren't waiting for us to figure this out. They're already using AI, and their expectations are changing.