🤖 AI-augmented negotiation
AI & decision support
AI-assisted negotiation: research on negotiating agents (N. Brown et al., CICERO, Science 2022), decision support and bias detection; preparation and debriefing with large language models.
Full detail in the “Origin & history” section below.
Author of the library
Negotiation lecturer at KEDGE Business School
A rare threefold background serving negotiation: teaching, commercial justice and audit, backed by experience as a company director in Bordeaux.
"Preparing Your Negotiation with AI" consists of enlisting a large language model (LLM) as a preparation copilot ahead of the table: you have it map out the interests of each party, generate and stress-test your BATNA (Best Alternative to a Negotiated Agreement), structure arguments and anticipated objections, then draw up a session plan. The technique does not replace the negotiator's strategy: it industrializes the preparation phase, historically the one most correlated with performance but the most neglected for lack of time. It turns an hour of solitary brainstorming into an adversarial simulation where the AI plays in turn the mirror, the opponent and the coach. Its value lies in the quality of the prompt and in the critical human eye that filters out the model's hallucinations and biases.
At a glance
Vigilance: moderate (4.0/10) · Preparation required: 8/10
Indicative profile: it situates the “Négociation augmentée par l'IA” family as the AI-augmented negotiation school practises it, not this technique taken in isolation. Techniques from the same family and school therefore share the same profile. NEGOCOACH editorial rating out of 10, non-experimental · the higher the “relational risk” value, the more costly the technique is to the relationship.
NEGOCOACH assessment
Tactical potential 5.7/10 (effectiveness, impact, discretion) and vigilance moderate (relational and ethical risk): two distinct readings, deliberately never merged into a single score that would reward risk. NEGOCOACH editorial rating calibrated from the “Négociation augmentée par l'IA” family and the “AI-augmented negotiation” school. Each criterion is rated out of 10; click to understand what it measures.
How far the technique can carry the negotiation in the intended direction when it is well executed.
Strength of the effect produced on the counterpart's perceptions, emotions and decisions.
How hard it is for the other party to notice the technique is being used. A high value = very discreet.
The information, analysis and rehearsal required upfront to use it effectively.
Potential cost to the relationship and to trust if the technique is spotted, refused or fails. A high value = riskier.
Moral acceptability: fairness, transparency and respect for the counterpart's autonomy. A high value = more defensible.
Documented school
The school this technique stems from is documented by recognised work and established practice, without experimental consensus. This indicator qualifies the school, not this technique taken in isolation.
Indicative NEGOCOACH editorial rating, for teaching purposes. For “Relational risk”, a high value signals a cost to the relationship, not a quality.
The technique has no single author: it is an emerging practice born from the meeting between the classic frameworks of principled negotiation (Roger Fisher and William Ury, "Getting to Yes", Harvard Negotiation Project, 1981, from which the notions of BATNA and underlying interests derive) and the massive spread of generative LLMs from late 2022 onward (ChatGPT). It was formalized and academically tested as early as 2024, notably by Tim Cummins and Keld Jensen ("Friend or foe? Artificial intelligence and negotiation", Journal of Strategic Contracting and Negotiation, 2024) and by work on AI-assisted humanitarian frontline negotiation (arXiv, 2024).
A structured preparation method that uses a language model as a cognitive partner in order to, before any real interaction: (1) inventory the parties' positions and deep interests, (2) formulate and evaluate one's own BATNA as well as the other party's presumed BATNA, (3) define one's ZOPA (zone of possible agreement) and reservation points, (4) produce a battery of prioritized arguments and anticipate objections with their counters, and (5) sequence a session plan (opening, anchoring, conditional concessions, closing). The AI works through iterations of prompts; the negotiator remains the sole decision-maker and verifies each output.
Application by context
Before a B2B sales meeting, the salesperson asks the LLM to simulate the buyer and their three most likely price objections, then to generate for each a value-based response and a differentiation argument, which they will keep only after critical filtering.
The buyer has the AI map out the supplier's probable cost structure, its concession levers (volume, payment terms, exclusivity) and its presumed BATNA, so as to enter the session with a numerical anchor and ready counterparts.
An HR director prepares a company-level negotiation by having the LLM list the unions' underlying interests beyond their stated wage demands (recognition, working conditions, job security) and the deadlock scenarios to defuse.
A crisis unit has the AI generate a decision tree of the possible demands of a hostage-taker or an attacker, with, for each branch, the stalling responses and the red lines never to cross, keeping the human as the sole master of the decision.
A ministerial adviser prepares a coalition negotiation by asking the model to model each party's breaking points, the bargaining chips (portfolios, reform timetable) and a realistic ZOPA on the common programme.
The buyer has the LLM estimate the market price range, the seller's probable motivations (relocation, inheritance, deadline) and a graduated concession plan linking price reduction, signing date and conditions precedent.
Before a negotiation with a Japanese partner, the negotiator asks the AI for a note on the expected codes (importance of nemawashi consensus, pace, the place of the unsaid) and reframes their arguments to avoid cultural missteps, while validating with someone who knows the ground.
In a tense estate division, an heir uses the LLM to distinguish positions ("I want the house") from each party's real interests (emotional attachment, financial security), and to explore creative options such as buying out the other heirs' share or shared usufruct.
Counter-techniques
Negotiation is also played on defence. Here is how to recognise this technique when it is used against you, and turn it around.
The signals that give it away
The counters that defuse it
Turn it into an advantage
Name the manoeuvre: said out loud, a technique loses most of its power.
Reacting emotionally instead of coming back to the facts.
Strengths: massive time savings on preparation; a mirror effect that reveals blind spots and argumentative weaknesses; an adversarial simulation available 24/7 and without judgement; creative broadening of the space of options; democratization of "expert-level" preparation for inexperienced negotiators; systematic structuring (interests, BATNA, ZOPA, plan). Weaknesses: risk of hallucinations and invented data (market figures, the opponent's BATNA) taken at face value; model bias (studies show variable and sometimes discriminatory salary recommendations depending on the profile); confidentiality, entering sensitive information into a consumer tool exposes you to leaks; the illusion of mastery and overconfidence; homogenization of strategies if all players use the same prompts; a dependence that atrophies the negotiator's judgement.
Particularly useful when the stakes are high but preparation time is limited; when you lack experience with the type of deal or the sector; when the negotiation is multi-party or multi-variable (many parameters to combine); in preparing cross-cultural negotiations where you are unfamiliar with the codes; to train beforehand through simulated role-plays. To avoid or strongly frame when the information is highly confidential (use of a private/local model is then required), when the human relationship and intuition take precedence over analysis, or in a crisis situation where only the human must decide.
Business · Cummins & Jensen: the controlled ChatGPT experiment at the table, In a study published in 2024 (Journal of Strategic Contracting and Negotiation), Tim Cummins and Keld Jensen had pairs negotiate under three configurations: with no AI on either side, then with only one of the two parties assisted by ChatGPT. The aim was to measure whether access to the model conferred an advantage on the negotiated outcome. The authors observe that AI changes the dynamic and can help with preparation and option generation, while stressing its limits (reliability, contextual judgement) and the need to keep the human in control. This work is one of the first to empirically document the effect of an LLM in a negotiation context.
Diplomatic · AI and humanitarian frontline negotiation (representative scenario), Research conducted in 2024 on humanitarian frontline negotiation (interviews with thirteen experienced negotiators, tests of ChatGPT-based tools) illustrates a realistic use: a negotiator preparing access to a conflict zone has the AI analyse the context, the interests of the armed actors and options for dialogue, keeping the decision strictly human. Practitioners see in it support for case analysis and creativity, but express strong reservations about confidentiality and model bias, hence the watchword "ChatGPT, don't tell me what to do". This case illustrates the recommended stance: the AI prepares, the human decides.
Facing an opponent who is themselves AI-prepared, the arguments often sound too smooth, symmetrical or "generic": spot the standardized vocabulary, the exhaustive lists of objections and the perfectly round anchors. Defence: ask specific and unexpected context questions that the AI could not have anticipated (operational details, the history of the relationship), move the discussion onto relational and emotional ground where the script holds up poorly, and test the real depth of your interlocutor's knowledge. Remember that the AI prepared scenarios, not the flexibility to improvise; step out of the expected script to regain the initiative.
Limits: uneven reliability (hallucinations), a lack of real understanding of the human stakes, potentially outdated training data, and an inability to grasp the non-verbal or field intuition. Ethics: the confidentiality of the data entered is a major issue (trade secrets, GDPR, third parties' personal data); the model's biases can reproduce discrimination (studies show differentiated recommendations according to presumed gender or origin); the asymmetry of access to AI can create inequity between parties; finally, having the AI play the opposing party or manipulate the preparation raises the question of negotiating fairness. Golden rule: the AI prepares, the human decides and takes responsibility.
Related techniques: classic preparation by matrix (interests/positions/BATNA/ZOPA) which it digitizes; role-play and negotiation simulation, here automated by the AI playing the opponent; "red teaming" or devil's advocate applied to one's own arguments; opposing-party analysis (stakeholder mapping); AI-assisted debriefing after the session; and, downstream, real-time AI assistance during the negotiation (live suggestions, bias detection), which extends preparation into action.
Quick exercise
Answer in your head, then reveal the solution. Memory is built through active recall.
Frequently asked questions
"Preparing Your Negotiation with AI" consists of enlisting a large language model (LLM) as a preparation copilot ahead of the table: you have it map out the interests of each party, generate and stress-test your BATNA (Best Alternative to a Negotiated Agreement), structure arguments and anticipated objections, then draw up a session plan. The technique does not replace the negotiator's strategy: it industrializes the preparation phase, historically the one most correlated with performance but the most neglected for lack of time. It turns an hour of solitary brainstorming into an adversarial simulation where the AI plays in turn the mirror, the opponent and the coach. Its value lies in the quality of the prompt and in the critical human eye that filters out the model's hallucinations and biases.
It sits on the line: effective, but it can tip into manipulation if it exploits an information asymmetry. Use it with measure and without deliberate deceit.
Reacting emotionally instead of coming back to the facts. The right reflex: slow down and reformulate.
NEGOCOACH does not assess the experimental validation of this technique in isolation. What we document is the grounding of its source school (AI-augmented negotiation): documented school. Full detail is in the "At a glance" section of this page.
Practise with AI
Copy, paste into your assistant, replace the [brackets]. Works with ChatGPT, Claude, Gemini, Mistral, Perplexity.
Build your plan before the meeting
You are an expert negotiation coach. Help me prepare to use the "Preparing Your Negotiation with AI" technique in the following situation: [describe your situation]. Give me: the conditions for success, a 3-step script, my counterpart's likely objections and how to answer them.
Rehearse against an AI counterpart
Play the role of my counterpart in a negotiation. I am going to test the "Preparing Your Negotiation with AI" technique. React realistically and with resistance, do not give in too quickly, then at the end analyse my performance and suggest 3 concrete improvements.
Analyse a past negotiation
Here is how my negotiation went: [paste the exchanges]. Analyse whether the "Preparing Your Negotiation with AI" technique was used well, what worked, the mistakes made, and spell out precisely what I could have done better.
References
Founding works of the 🤖 AI-augmented negotiation school this technique belongs to.
Human-level play in the game of Diplomacy (CICERO), Science
ArticleFAIR (Meta) - N. Brown et al. · 2022
Noise: A Flaw in Human Judgment
BookD. Kahneman, O. Sibony & C. Sunstein · 2021
Co-Intelligence: Living and Working with AI
BookE. Mollick · 2024
AI-assisted negotiation: research on negotiating agents (N. Brown et al., CICERO, Science 2022), decision support and bias detection; preparation and debriefing with large language models.
On video
Videos to picture Preparing Your Negotiation with AI and anchor it through examples.
A verified video selection is being enriched; the search above already surfaces the best videos on the topic.
Technique map
Every technique sits within a network: what it draws on, what it combines with, where it applies, and how to defend against it.
Spot its signals, neutralise it and turn it around with the defensive playbook on this page.
See the counter-techniques"Preparing Your Negotiation with AI" consists of enlisting a large language model (LLM) as a preparation copilot ahead of the table: you have it map out the interests of each party, generate and stress-test your BATNA (Best Alternative to a Negotiated Agreement), structure arguments and anticipated objections, then draw up a session plan. The technique does not replace the negotiator's strategy: it industrializes the preparation phase, historically the one most correlated with performance but the most neglected for lack of time. It turns an hour of solitary brainstorming into an adversarial simulation where the AI plays in turn the mirror, the opponent and the coach. Its value lies in the quality of the prompt and in the critical human eye that filters out the model's hallucinations and biases.
Name the manoeuvre: said out loud, a technique loses most of its power.
Reacting emotionally instead of coming back to the facts.
Our programmes turn theory into a concrete advantage.