Negotiations 2025: What would it mean to negotiate with an A.I.

April 19, 2018

Soon-to-become reality or pure fiction?

Ten years ago the situation described in "Good Morning Michael" might have sounded like pure SciFi. But with today’s fast paced development in robotics and artificial intelligence, the scenario is becoming very much likely. The development began with the first computers and then with the internet. Since then we see data based management and decision making becoming state of the art in business. All companies benefited from this development. Nowadays there is data available on almost every relevant business topic, so why not to use it for negotiations too?!

Some people think that artificial intelligence will not be used in high-stakes transactions, i.e. when two parties are negotiating deals with high financial impact. But exactly these kinds of situations require a negotiation style which minimizes the risks and the mistakes. And humans, as emotional beings, are predestined for making mistakes because they get emotional.

In addition, almost all deals can be numbered. Every negotiation problem can be represented in figures. As soon as the negotiating parties have slightly different preferences on the negotiation topics, you can quantify them. And here is the fact: machines, especially artificial intelligence or intelligent robots, are much better at dealing with numbers than humans are. They are faster and more efficient.

Lastly, if you have a machine which is learning, and which is getting better with every interaction over time, you as a company will have the perfect employee: never getting sick, never asking for a salary, and always giving its best. It’s a low-cost high-gains situation for every corporation.

These four reasons will be why the usage of A.I. for negotiations in companies will become state of the art sooner than later. In the beginning, the A.I. will be used for a better preparation of the humans negotiating with each other. Then some corporations will replace human negotiators entirely with A.I. There will be only one human to final-check the deal, as showed in the example above. These early movers will gain from the human-machine interaction. Even if the A.I. will seem indeed "artificial“ in the beginning, there will be a third step; empathy coaches, who will teach the A.I. conflict resolution skills and ways to overcome disputes easily in negotiations. With the A.I. getting better in time, it will get full decision-making authority, without a human cross-check. Human interaction will be only needed when or if something goes wrong in the implementation. A forensic data-analyst will be able to analyze what happened and to train the machine to avoid making this mistake in future. More and more companies will see that the usage of A.I. in negotiations will work out well. They will implement their own machines. In the end, huge commercial deals will be negotiated only between artificial intelligences.

If this is to come, how do we deal with it?

Will we lose our humanity? Maybe. The first reaction will consist of fear and the wish to defeat this development. But we as humans will have to learn to deal with it. That will be the only constructive strategy possible. It also depends on which rules the A.I. will have programmed. If a negotiation A.I. always has to seek mutual benefit, and not only for the corporation it represents, we could see a huge increase of high quality deals worldwide. Both parties at the bargaining table could benefit from that.

How the use of A.I. could improve negotiations

Artificial intelligence could be used in negotiations in various ways. First will be the usage of A.I. to improve and to speed up one's own preparation. The availability of a

large amount of data, interconnecting the information in an intelligent way will provide the opportunity for an extensive, holistic preparation. In addition, A.I. could be used by one or both parties during the actual interaction to support their own arguments with standards (facts, figures, studies, etc.). A Negotiation A.I. could furthermore negotiate with multiple parties simultaneously, which would lead to a significant reduction of interaction time. The last way to create benefit through the usage of artificial intelligence in negotiations would be the increase in simplicity of documentation. Even the entire in-person negotiation sessions could be recorded and stored for documentation purpose. All exchanged documents, files, notes, e-mails and so forth could be automatically preserved by the artificial intelligence. This would lead to an overall increase of negotiation knowledge in the whole company.

In Negotiations of Commercial Deals, i.e. Purchasing Dept. vs. Sales Dept.

Let’s stay at the example mentioned above, the commercial interaction between the purchasing department of company A and the sales department of company B. Both companies annually conduct several rounds of year end negotiations, where they fix terms of businesses, like savings, costs, delivery, quality, etc.

If A’s purchasing and B’s sales department are using A.I. to prepare, the system will provide a tremendous amount of information. It will be able to combine these and to learn from the past interactions and deals. It will be effortless to prepare information like:

  • data on past deals
  • present and past preferences
  • current performance KPIs
  • production cycles
  • past and current quality incidents
  • past and current logistics issues
  • product lifecycle of the whole portfolio
  • benchmarks (e.g. price development of raw materials)
  • research, studies, publications
  • similar deals in the same or in a different industry
  • local regulation and laws

This data would make one's own preparation more efficient. But even more, it could be used to make assumptions about the other company’s wishes and needs. Combined with

  • analysis of news feeds on the company (successes, issues, scandals), and
  • information on the human counterparts from social sites for profiling (analyzing online personality)

the A.I. could be used to prepare not only on the factual side, but also on the persons and on the process. If the system is used in the first step just to help the human prepare then it could learn essential questions (checklist) to ask its human colleague.

Even if the A.I. will someday replace one negotiation counterpart, let’s say the purchasing people, there will be a so-called pre-negotiation of a deal in an A.I.-human interaction. The higher the financial stakes of the deal, the more probable that a human will be the final decision making instance.

The potential downsides are still to be analyzed 

The machine/A.I. and human interaction is a focal point of research in many universities and companies worldwide. Designed to bring benefit to the humans, (deep) A.I. can nevertheless be perceived as a threat by humans, especially in negotiation situations. It might happen that people unconsciously show a hostile attitude toward the A.I. Some of them might feel automatically inferior and in a weak bargaining position. Some others might want to compete with the A.I. Even if the A.I. proposes a mutual-gains solutions, it might happen that humans would be nevertheless angry because of interacting with a non-human counterpart and showing no willingness to negotiate with the A.I. again at a later date.

ISMAN & Partner has recently started a research project on the topic „What a human - A.I. negotiation would look like?“. The research aims als to find answers to questions like

  • Are humans, if their emotions switch into negative still able to trust the A.I.’s proposal and to be creative in the common solution finding?
  • Will people use or try to use distributive tactics? How will the participants react when they play negative emotions and the A.I. “reads“ (and verbalizes it) that the emotions are only acted and not real?
  • How would humans experience negative emotions against the A.I.? Would they become angry and will they desire to hurt it and break up the negotiation, rather than satisfying themselves?
  • Will humans have the feeling that there are options for controlling the situation (secondary appraisal process)?
  • What goals do humans set themselves before the negotiation starts? And will they follow their goals based on the perceived emotions?

The results of the research shall be used to train humans in negotiating with artificial intelligence, as well as to ensure that the work of machines is both effective and responsible — that it is fair, transparent, and audit compliant. If humans have no fear negotiating with machines, and if machines are trained to follow a mutual gains approach, we’ll be able to see an overall increase in negotiation quality worldwide.

ISMAN & Partner
ISMAN & Partner ist eine Unternehmensberatung, die nationale und internationale Konzerne, mittelständische Betriebe und Start-ups, Organisationen und Institutionen bei komplexen Verhandlungs- und Konfliktlösungsprozessen begleitet. 2015 von Calin-Mihai Isman gegründet, unterstützen die Experten für Negotiation & Mediation Manager und Mitarbeiter aus den Bereichen Sales, Einkauf, M&A, Contracting, HR oder IT.


Bleiben wir in Verbindung.

Mit dem Absenden dieses Formulars bestätigen Sie, dass Sie unsere Datenschutzerklärung zur Kenntnis genommen haben und dass Sie mit der Speicherung Ihrer eingegebenen Daten einverstanden sind.
Vielen Dank!
Wir haben Ihre Nachricht erhalten.
Entschuldigen Sie. Etwas ist schief gelaufen. Probieren Sie es doch bitte noch einmal.


Keep in Touch.

By sending this form you confirm, that you accept our Privacy Policy and that you agree to the storing of the information inserted!
If you send us requests via our contact form, the information you insert including the contact data you insert will be stored for the purpose of the processing of your inquiry and for the case of follow-up questions.

We do not pass on data on to third parties without your explicit approval. The processing of the inserted information takes place therefore on the behalf of your consent (Art. 6 Abs. 1 lit. A DSGVO). You can revoke your approval at any time. To do this you can send us an informal E-Mail. The lawfulness of the data processing until revoked remains unaffected by the revoke. 

The Information you inserted in the contact form, will be stored by us until you request deletion, revoke your approval for storing the data or the purpose of storing the data is no longer required (e.g. after closing the process of handling your inquiry). Mandatory legal provisions – especially storage periods – remain unaffected.

For more Information please read our Privacy Policy.
Vielen Dank!
Wir haben Ihre Nachricht erhalten.
Entschuldigen Sie. Etwas ist schief gelaufen. Probieren Sie es doch bitte noch einmal.

Interested in learning more?

Möchten Sie Mehr über uns erfahren?