AI is helping agents sound more knowledgeable than they are 

Thinking big about residential real estate success requires a big-picture perspective. Industry Decoded features industry experts who can enrich your understanding of issues affecting the industry as a whole.

The views expressed in this column are solely those of the author.

I've used AI to run comps and watched it get things wrong, delivering incorrect information in the same confident voice it uses when it's right. If I didn't know the market, I might have handed that number to a seller and called it advice.

I use this technology every day, so nobody needs to talk to me about embracing change. It runs the parts of my business my brain isn't as good at managing. But I've sold more than 500 homes, and that's how I know when it's wrong. A lot of the agents leaning on AI the hardest don't have that experience.

It's easy to sound smart

AI chatbots can give convincing answers to any number of real estate questions. That becomes a problem when an agent who can't spot the inaccuracies sounds as confident as the agent who can, and the person signing the contract can't tell them apart.

In June, NAR's board took up a proposal that would have required Realtors to disclose when they lack "the knowledge, information, or skills about a property type or geographical area to protect and promote the interests of their clients considering the complexity and particularities of the transaction." The Executive Committee recommended defeating it. The board of directors voted 497 to 356 to send it back to the Professional Standards Committee, which had already been working on it for two years.

Some of the objections were fair. The language was vague, and new agents have to learn somehow. It takes almost nothing to get a license, and I say that as someone who walked through the same door in 2014.

How agents are actually losing consumer trust 

Despite those caveats, there's one argument I can't let go of. Real Estate News reported that during the board meeting, a director from Florida said that once an agent discloses that kind of knowledge gap, "you've lost that trust with the consumer."

That trust is already waning, with or without an NAR disclosure rule. Spend five minutes on Facebook and you'll see agent content that is plainly fake, and consumers can spot it as fast as we can. 

A client doesn't lose faith in the agent who says, "I haven't sold a farm, so I'm bringing in someone who has." A client loses faith after finding out the agent never had the knowledge or skills to adequately promote their interests — and by then, it's usually too late to change the outcome.

AI is obscuring knowledge gaps, not erasing them

NAR's latest technology report says 23% of Realtors use AI daily, and other surveys suggest AI usage among agents may be much higher. Good. What worries me is how many treat it like the Bible: The machine said it, so it must be true. They paste it, post it and send it without checking a word.

AI can look at a listing photo and tell you the kitchen has been updated. It has never stood in that kitchen or noticed the floor sloping toward the back wall, and it can't tell you what else it doesn't know or if its interpretation is wrong.

Clients used to have clumsy ways of spotting an agent who was in over their head: a vague answer, a long pause, a phone call that never got returned. Those signals were never foolproof, and AI is erasing them. An agent who closed two deals last year can now publish a market analysis every morning and answer a land question in flawless paragraphs. 

They look smarter and busier than they are, and intentionally or not, they are misleading people. Every client who eventually figures that out trusts the rest of us a little less.

AI is a resource, but agents are the ones held accountable for mistakes

Peter Parker's Uncle Ben had it right: With great power comes great responsibility. AI doesn't hold a license. We do. I serve on the Ohio Realtors grievance and mediation panels, and every name on a complaint belongs to a person. NAR's own AI policy kit for brokers says agents and staff are "fully accountable" for the AI content they use. Article 1 of our Code of Ethics already tells us to protect and promote our clients' interests. That duty doesn't shrink because AI produced a draft more quickly.

NAR's directors met again in New Orleans last week. While one session focused on the risks of using AI and the importance of maintaining consumer trust and compliance, the knowledge gap proposal wasn't a topic of discussion.

I hope that changes. I don't know if that proposal's wording is right, and I'd want it fixed before panels like mine ever had to apply it. But requiring agents to be honest about what they don't know is not what's putting consumer trust at risk. Just read what some agents are publishing under their names and the Realtor logo.

The standard I'd hold all of us to is simple: Use AI as a tool, then check its outputs against a reliable source. If you can't explain the answer or data on your own, then you don't know enough to share it with your client — and they deserve to hear that from you first.

Nathan White is a Central Ohio Realtor and ABR Instructor. He has served as Chair of the Columbus Realtors MLS Committee and has been active in professional standards, mediation, ombudsman work and industry education at the local and state levels.

Nathan is passionate about protecting consumers, strengthening the role of the MLS, and helping agents build better businesses through innovation, transparency and accountability.