Many of the questions I’m getting from agents these days are about how to deal with AI-generated advice.

“My seller says ChatGPT thinks I should list higher.”

“My buyers sent me an AI-generated demand for a $40,000 reduction after the inspection. They want me to forward it to the listing agent and I know it will not go well.”

“The other agent sent three pages explaining why their offer is fair. Beautifully written. Almost entirely disconnected from the property we’re negotiating.”

Does this sound familiar?

There is now another voice in the negotiation. It is available at midnight, remarkably patient and very good at sounding certain. It doesn’t need to book a showing or have an uncomfortable conversation with the seller.

We do.

I could write about why AI gets things wrong. I’m more interested in what its presence requires us to become better at.

 

The advice is only part of the disruption

 

A recent study, Algorithmic Disruptions to Expertise, examined how 15 real estate professionals in North Carolina experienced Zillow’s Zestimate. The researchers found that automated valuations were changing how clients evaluated their agents’ advice. The algorithm had become a competing source of authority.

It’s one tool in one U.S. market. But the pattern will be familiar to Canadian agents: clients using an algorithmic answer as the benchmark against which they judge your expertise.

You recommend one approach. Their AI recommends another. Now they need to decide who to believe, and that changes the work.

Having information is useful. Helping someone understand what it means, what is missing and how it applies to their circumstances is increasingly where we demonstrate our value.

 

A convincing argument is not a negotiation strategy

 

We have always underestimated the depth of negotiation in real estate. Consumers do it. Agents do it too.

If negotiation means arguing for a lower price, AI can give you an impressive argument. But the pathway to that lower price might involve understanding the seller’s needs, offering a different closing date, assessing their alternatives or allowing them to accept less without feeling defeated.

When the prompt is simply “How do I get this house for less?” AI is being asked to recommend a strategy before it understands the negotiation.

A skilled negotiator discovers information that changes the strategy.

What matters to these people? What are they afraid of? What happens if we don’t reach an agreement? Is this demand a firm boundary, an opening position or an attempt to feel more in control?

AI can help explore these questions. But someone needs to recognize that they need asking.

 

They hired you. Do they trust you?

 

This is where things get uncomfortable.

A client can like you, hire you and still doubt your judgment. They can also respect your expertise while wondering whether your recommendation serves their interests or your desire to get the transaction completed.

Professional trust has both a head and a heart.

Head trust asks, “Do you understand this? Can I rely on your judgment?”

Heart trust asks, “Do you care about what happens to me? Will you put my interests first?”

I hear a version of this story from agents over and over. The CMA was accurate and complete. The agent recommended a strategy for pricing, timing and marketing. The seller chose a higher price and said, “Let’s try it for three weeks.”

Three weeks later there are no offers, few showings and a seller reluctant to reduce.

Maybe the seller was never convinced by the numbers. Maybe they suspected the agent cared more about a quick sale than the best outcome. One is a head trust problem. The other is a heart trust problem. They need different responses.

When a client rejects your recommendation, consider what might be missing before producing another six comparable sales. More evidence may help. But it won’t necessarily address a concern about your intentions or a fear the client hasn’t expressed.

Meanwhile, a polished AI response can look very competent, and its patient tone can feel reassuring.

Our experience matters, but we need to make our reasoning visible. We need to explain uncertainty, explore alternatives and demonstrate that we understand what the decision means to the person making it.

 

Get curious before you get defensive

 

When someone brings AI advice into a conversation, there is an opportunity to learn.

What did they ask? What information did they provide? What about the answer felt useful or convincing? What outcome are they hoping it will help them achieve?

You don’t have to agree with the recommendation to respect their effort to make a good decision.

Then examine it together. Add missing context. Test assumptions, including your own. Explore the consequences of following the advice.

Occasionally, AI may have identified something you missed. Being willing to recognize that is part of earning trust.

The objective is a better decision.

 

Where this leaves us

 

AI raises the bar for agents. It also makes the depth of skilled negotiation easier to see.

On Nov. 24, I’m partnering with REM to teach Negotiation Intelligence in the Age of AI, a 2.5-hour course exploring professional trust, negotiating power, human behaviour and practical ways to navigate AI-influenced conversations.

Participants will also receive three months of access to NIA, our Negotiation Intelligence Assistant, for support with preparation, strategy and difficult conversations.

Yes, she is available at midnight too.