How AI is Reshaping Negotiations, Part 2

In Part 1 of this series, I introduced the concept of AI as an exoskeleton to the skilled negotiator, a tool that enhances preparation and analysis and leaves judgment and empathy to the human. That article explained that the best negotiators have created four practices: auditing their presence online, over-communicating on screen, pre-meeting stress-testing with AI, and making time in-person to do what is important. It had brought in more response than anything I had previously written, and the greater part was reduced to a single question:

yes, but where does it end?

That question deserves its own article because the honest answer is that we are still finding out.

In a 2024 paper published in SAGE’s Negotiation Journal, Falcão Filho explored the maturing relationship between AI and negotiation and found that we are quickly transitioning from AI as a support tool to AI as an active participant in negotiation strategy.¹ Additionally, a study by MIT and Harvard Law School Program on Negotiation in 2024 has demonstrated that coaching bots, AI systems designed to challenge your negotiation logic before you enter a live session, can help negotiators to achieve better results by as much as 60 percent.² That is not a marginal gain. That is a structural advantage.

The Persona Revolution

What most people still think of as AI in negotiations is basic: summarising meeting notes, drafting follow-up emails, pulling competitor data quickly. All that is actual and effective. However, the more advanced use that emerged consistently in my doctoral work was the AI personas to stress-test strategy prior to anyone being seated at the virtual table. This is right in Element 6 of my Business Negotiation Framework that I have created based on 16 interviews.

One participant described his team’s approach:

“We use AI personas to simulate a CFO or procurement manager. We see how they might push back, what arguments hold.”

Another participant captured why this changes the preparation game:

“We used to rely on intuition. Now AI tells us what worked before and what didn’t. It’s like having a coach who remembers every deal you’ve ever done.”

The research on the MIT coaching bot proves what my participants were already practicing. Their most effective tools are in the preparation stage, where AI can assimilate your proposal, simulate your opponent, and uncover the flaws of your reasoning before they turn into flaws of your negotiation. For the full picture of what systematic preparation looks like in practice, see “From Gut Feel to Data-Driven”.

The Ethical Tensions Nobody Is Talking About Enough

Here is where the conversation gets uncomfortable. And where I think we need to be more honest as a profession. Almost all of the interviewees in my doctureal study who frequently utilized AI were also very concerned about it. Not effectiveness but authenticity, transparency, and gradual loss of human judgment. These are directly related to the issues of trust that I have discussed in “Why Trust Is Harder to Build on a Screen”: AI that is abused or concealed from your counterpart can be more damaging to cognitive trust than it helps build it.

One participant explained it simply:

“Transparency matters. If you use AI for insights, fine, but don’t pretend you came up with it.”.

One of the participants brought up an issue which should be viewed with significantly greater concern:

“Younger professionals risk losing the instinct. AI gives answers too fast. Negotiation is about judgment, not autocomplete.”

In March 2025, the Harvard AI Negotiation Summit, co-hosted by MIT and Harvard Law School, convened a group of top scholars to discuss precisely this tension.³ Whether AI can enhance the outcomes of negotiations was not their main issue, the data indicates they can, in preparation. Their question was how to preserve human autonomy in a process that is becoming increasingly mediated by algorithms that optimise the result, but which cannot consider the relationship, ethics, and long-term cost of a good deal that looks good on a spreadsheet.⁴

Three Positions the Best Negotiators Hold Simultaneously

First, they aggressively apply AI during the preparation process and not in live negotiation. The preparation is where AI adds the most value without the most risk. It is one thing to use AI to create meeting briefs, scenario analysis, and red-team challenges prior to a negotiation and quite another to use AI cues in the middle of a negotiation when the output cannot be verified, and the relational consequences of error are immediate.

Second, they are transparent with their counterparts concerning the purpose of AI in their process. This is not weakness; it is credibility. Participants in my research consistently reported that over-use of AI that was later discovered damaged trust more than the AI itself helped. One participant reflected: “AI can prepare you, but it can’t care. Negotiation is still about human connection.”

Third, they consider AI literacy a novel negotiation skill, not a substitute for the previous ones. The ability to understand when to follow the advice of an AI and when to ignore it will be a conscious practice, sincere self-reflection, and the readiness to be wrong in the preparation room in order to be right in the negotiation room.

Whether AI should be a part of negotiations is not the question. It does. The question is whether we are thoughtful enough about where it belongs and where it does not.

Thanks again, Tim H.

ps. the picture is from last week snowboarding with my brother.

@misc{herglotz2026dn06,
  author       = {Herglotz, Tim},
  title        = {{How AI is Reshaping Negotiations, Part 2}},
  year         = {2026},
  month        = {4},
  number       = {6},
  howpublished = {Tim Herglotz, entry № 06},
  url          = {https://timherglotz.com/ai-b2b-negotiations-ethical-frontier/}
}