How to Get Real Estate Agents to Actually Use AI Tools

By Emily Terrell — Top Coach and Speaker at Tom Ferry International. Real estate’s leading voice on AI, systems, and social media.

To get real estate agents to use AI, diagnose which resistance you’re facing before you demo anything. Agents burned by bad output need to rebuild on their own listing. Agents protecting their identity as relationship-driven need peer proof, not a coach’s demo. This guide covers both diagnoses and the rollout that follows each one.

Key Takeaways

• Most “skeptical” agents already tried AI, got generic output, and drew a rational conclusion — arguing them out of a correct conclusion doesn’t work.
• The burned skeptic and the identity skeptic need opposite moves, and treating them as one group is why rollouts stall.
• Compliance fear is the objection agents don’t say out loud, and naming it first does more than any capability demo.
• Give one prompt for one task, not a library — a library is a decision, and a decision from a skeptic is a delay.
• Convert your top producers first and let peer pressure work downward, because coach proof reads as a pitch and peer proof reads as evidence.

What is AI resistance in real estate?

AI resistance in real estate is the gap between an agent having access to AI tools and choosing to use them on real work. It’s rarely a philosophical objection to technology. It’s a conclusion an agent reached after a bad first experience, or a defense of the part of their business they believe AI threatens.

That distinction changes the entire response. You don’t argue someone out of a conclusion they reached from evidence. You give them different evidence.

Why this matters for real estate agents

The tools are already in the building. According to NAR’s 2025 Technology Survey, released September 18, 2025, two out of three agents either agree (38%) or strongly agree (29%) that their brokerage provides all the tech tools they need. Access is not the constraint. Nobody is waiting on procurement.

The barriers agents actually name are smaller and more fixable than the ones leaders assume. In RPR’s February 2026 survey of 225 NAR members, reported by HousingWire, the most common barriers to using AI more regularly were not enough training (16.82%), too many tools to choose from (15.91%), and not being sure where to start (12.73%). That same survey found only 8% of respondents said they don’t use AI and never plan to.

Read those together. Almost nobody is a true refuser. The population you’re calling skeptical is mostly agents who don’t know which tool to open or what to do first.

“Agents are not resisting AI. They’re using it where it adds clear value.” — Reggie Nicolay, RPR

Which kind of skeptic are you dealing with?

Three patterns, and they need different responses. Diagnosing wrong costs you the conversation.

The burned skeptic

This agent tried it. They asked a chatbot to write a listing description, got something that read like a brochure from 2011, and quit. Their skepticism is correct given their inputs — the failure was the prompt, not the technology, but nothing in their experience told them that.

Demos make this agent worse, because a demo is you succeeding at the thing they failed at. What works is rebuilding the same failure in front of them, on their actual listing, with their actual MLS data. Not a sample file. Their address. They need to watch the prompt be the variable.

Expect visible frustration during that rebuild. That’s the segment working, not failing.

The identity skeptic

This agent says some version of “my business is relationships.” They’re not worried the output will be bad. They’re worried it will be good, and that being good makes the thing they’re proud of replaceable.

Do not demo to this agent. A demo confirms the fear. Put them next to a peer at or slightly above their production level who closed something using it, and get out of the conversation. Coach proof reads as a sales pitch. Peer proof reads as evidence.

The RPR data supports the format: respondents named hands-on workshops (57%) and use-case training tied to real tasks like CMA creation (56%) among the most helpful forms of AI training, second and third behind short video tutorials (69%).

The compliance skeptic

This is the one nobody says out loud, and it’s the most common. Agents worry that AI-generated marketing will publish something that costs them their license.

That fear is not irrational. RPR’s February 2026 survey found 63% of respondents cited accuracy of outputs as their top concern, followed by compliance or legal issues at 49% and fair housing concerns at 28%. And every REALTOR® has sat through the training that makes the exposure feel real — NAR requires Fair Housing and anti-bias training on joining and every three years after, on the same cycle as Code of Ethics, with the current requirement running from January 1, 2025 to a December 31, 2027 deadline.

Name the guardrails before anyone asks: the review step, the protected-class language rule, and the broker identification requirement on advertising. Removing the unspoken fear moves more agents than any capability demo. The specifics are in what actually risks your license when you use AI.

This is general information, not legal advice. Consult your broker and an attorney licensed in your state.

What actually converts each one

Four moves, in order.

Lead with the task, not the tool. Nobody wants AI. They want the follow-up they haven’t touched in three weeks off their plate. Start with the thing the agent already hates and already does badly, and don’t say the word “AI” in the room. It’s a listing description draft. It’s a follow-up script. The label is what they’re arguing with.

Give one prompt, not a library. A prompt library is a decision, and a decision from a skeptic is a delay. One prompt, one task, and no second prompt until the first one has been run on live work.

Measure in minutes they generated themselves. Ask how long the task took last time, before anything opens. Then run the clock. A skeptic doesn’t convert on your enthusiasm. They convert on their own number.

Send the output somewhere. A draft that lives in a chat window evaporates by Thursday. It goes into the CRM template, the saved prompt file, the email sequence — or the agent has a faster way to produce work they still don’t use.

How I use this in my own business

I close 70+ transactions a year in San Antonio on roughly five hours of active management per week, and I’ve lost this argument in my own office more than once.

The version that finally worked wasn’t training. I stopped trying to convince the agents who were pushing back and built the workflow with the two people already producing at the top. Six weeks later, the holdouts came to them, not to me — and they came asking a specific question about a specific task, which is a much better starting point than a general objection to technology.

The other thing I changed: I stopped opening with capability. Now the compliance guardrails come first, in the first five minutes, before anyone sees a single output. It cost me applause. It gets me agents who actually publish what they build.

Common mistakes

1. Treating all skepticism as one objection. The burned agent and the identity agent need opposite moves. One response for both converts neither.
2. Demoing to the identity skeptic. Every impressive output confirms the exact fear driving the resistance.
3. Handing out a prompt library. More options is more decisions, and skeptics resolve decisions by doing nothing.
4. Saving compliance for the end. By then the agent has spent the whole session quietly deciding not to publish anything.
5. Working the skeptics instead of the producers. The slowest path to a resistant agent runs through you. The fastest one runs through the peer one tier above them.
6. Declaring success at enthusiasm. Room energy is not adoption. A named workflow running 30 days later is adoption.

Frequently Asked Questions

Why do real estate agents resist AI tools?

Most don’t resist it philosophically. They tried it once, got generic output, and concluded it doesn’t work for their business — a reasonable conclusion from that evidence. Others are defending their identity as relationship-driven professionals. RPR’s February 2026 survey found the top barriers were insufficient training, too many tools, and not knowing where to start.

How do you convince a skeptical agent that AI works?

Stop convincing and start rebuilding. Run the task they already failed at, using their own active listing and their own MLS data, and let them watch the prompt change the output. Their frustration during that rebuild is the point. Demonstrations of your success at their failure tend to deepen resistance rather than reduce it.

Should I train my whole team on AI at once?

No. Start with your top producers and let adoption move downward through peer proof. Agents discount a coach or team leader recommending a tool and weight a peer at their production level heavily. Team-wide rollouts also surface all three resistance types simultaneously, which makes it impossible to respond to any of them correctly.

What’s the biggest hidden objection to AI adoption?

Compliance. Agents worry AI-generated marketing will publish something that costs them their license, and they rarely say it out loud. RPR’s February 2026 survey found 49% cited compliance or legal issues as a concern and 28% named fair housing specifically. Addressing the guardrails before anyone asks removes more friction than demonstrating another capability.

How many AI tools should an agent start with?

One tool and one prompt, applied to one task they already do badly. Prompt libraries and multi-tool comparisons create decisions, and a skeptical agent resolves decisions by postponing them. Add the second workflow only after the first one has run on live business for several weeks.

How do you measure whether AI adoption actually worked?

Count agents running a named workflow 30 days after the session, not survey scores or room energy. Capture each agent’s baseline hours on the task before anything opens, so the comparison is a number they produced themselves. Self-generated numbers move skeptics; borrowed benchmarks don’t.

What should a team leader do before running AI training?

Diagnose the resistance in the room and require real material as pre-work. Without an active listing or live lead in front of them, agents watch instead of build. The full agenda, timing, and pre-work requirements are covered in what AI training for real estate agents must cover.

Bring this to your team or event

Emily Terrell speaks at brokerage events, real estate conferences, and team trainings on AI, systems, and social media — the exact playbook in this post, delivered live to your audience. As a Top Coach and Speaker at Tom Ferry International and an active agent closing 70+ transactions a year, Emily speaks from the stage about what’s working right now, not theory. Recent stages include NAHREP and eXp Con. See keynote and training formats.

Book Emily to speak at your next event:
Email: eterrell@yourcoach.com
Phone: (210) 400-9191
Web: coachemilyterrell.com

For real estate agents who want to implement this: Get the weekly real estate prompt library at weeklyrealestateprompts.com or follow @coachemilyterrell on Instagram for daily systems and AI breakdowns.