.png)
In short: In a 2026 experiment, people agreed to higher rents when they thought they were negotiating with a human—even though their counterpart was AI. For procurement teams, the finding raises a practical question: how do you separate feeling good about a negotiation from getting a good deal?
Some participants in a rent negotiation experiment came away happier with the interaction, yet agreed to pay more. Everyone negotiated with the same AI system. The difference was that one group believed there was a person on the other side. [1]
There is something familiar here for anyone who works in procurement. After a meeting, it can be easier to recall how accommodating a supplier seemed than to assess the terms you accepted. The research raises a further question: how much are our decisions shaped by who we believe we are dealing with?
In a study published in October 2026, Alexandra Mislin, Daniel Druckman and Louise Serre asked 205 participants to negotiate the rent for an apartment. All participants dealt with the same intermediary system, powered by a large language model. One group was told they were negotiating with a human; the other was told they were dealing with AI.
Those who believed their counterpart was human felt better about the relationship but agreed to a higher final rent. The underlying system was identical in both groups. What differed was the participants’ understanding of who—or what—they were negotiating with. [1]
This does not show that AI is a better negotiator than a person, nor can the findings be assumed to apply to corporate purchasing. It does show that, in this experiment, greater satisfaction with the relationship went hand in hand with a higher price.
Imagine meeting a potential maintenance provider for the first time. They listen carefully, understand the challenges at your plant and seem easy to work with. Their quote is a little higher, but you leave feeling that they would be there for you if something went wrong.
That may prove to be true. At this stage, though, you have only their proposal and the experience of one meeting to go on.
Now consider a provider that has met its deadlines for years, answered calls during unexpected shutdowns and dealt with complaints effectively. Your confidence in that supplier rests on performance you have seen. With the new provider, their approach during the meeting gives you a reason to expect good service, but that expectation has yet to be tested.
The distinction can easily disappear behind a phrase such as “They’re good to work with.” When that judgment influences a purchasing decision, it helps to be explicit about what supports it: a track record, references you have checked or the impression someone made in an initial conversation.
A proven supplier relationship has business value. A positive negotiation experience should be assessed separately.
The study suggests a useful exercise for procurement teams. After a negotiation, write down the substance of the offer: prices, deadlines, commitments and anything still unresolved. Then ask yourself: would I judge these terms the same way if I were told an AI system had proposed them?
If the answer is no, consider what is making the difference. Is it the supplier’s past performance? A clear, specific explanation of how they would handle a problem? Or the fact that you felt listened to and understood?
In the maintenance example, a higher fee could be justified by a shorter time to restore service, guaranteed capacity or verified experience with your equipment. When those benefits are set out in the proposal and can be checked, there is a concrete basis for paying more. If the case rests mainly on a good first impression, follow-up questions and reference checks can help establish whether that confidence is warranted.
This is a practical suggestion inspired by the findings, not a technique tested in the study. Its purpose is to help identify what, beyond the terms on the table, is shaping the decision.
Companies using AI in supplier negotiations need to look at both the terms they secure and the supplier’s experience of the process. A favourable price is one consideration. It also matters whether the supplier understands what they have committed to, can ask for clarification and knows whom to contact afterwards.
For procurement professionals, the broader lesson is to make sure that confidence in a deal is supported by evidence. Which benefits are written into the agreement? Which have been demonstrated through experience? And which are still expectations formed during the negotiation?
Does the study show that AI negotiates better than humans?
No. Both groups negotiated with the same AI system. The experiment showed that believing the counterpart was human or AI affected the outcome in that particular setting.
Is paying more for a better supplier relationship a mistake?
Not necessarily. More reliable delivery, available capacity or less risk when switching suppliers may justify a higher price. To assess the decision properly, the evaluation needs to account for both the expected benefits and the additional cost.
How can you tell a good impression apart from the value of a supplier relationship?
A good impression comes from the interaction itself. The value of a supplier relationship can also be supported by evidence such as past performance, proven capabilities, available capacity or other specific benefits. When explaining a decision, make clear what your assessment rests on.