Personality Engineering with AI Agents: A New Methodology for Negotiation Research

Jared R. Curhan; Michelle A. Vaccaro

arxiv: 2605.20554 · v1 · pith:3BKCBGSQnew · submitted 2026-05-19 · 💻 cs.AI · cs.HC· cs.SI

Personality Engineering with AI Agents: A New Methodology for Negotiation Research

Michelle A. Vaccaro , Jared R. Curhan This is my paper

Pith reviewed 2026-05-21 06:18 UTC · model grok-4.3

classification 💻 cs.AI cs.HCcs.SI

keywords personality engineeringAI agentsnegotiation researchinterpersonal circumplexwarmth and dominance

0 comments

The pith

AI agents with engineered personalities enable rigorous tests of negotiation theory.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

This paper proposes personality engineering as a new methodology that uses AI agents to precisely parameterize, manipulate, and evaluate negotiator personality. Canonical negotiation theory requires balancing empathy against assertiveness and concern for the other against concern for the self, yet people find these tensions difficult to sustain consistently. AI agents avoid those human limitations through precision, consistency, and scalability. The approach adopts the interpersonal circumplex, with its dimensions of warmth and dominance, as the coordinate system for controlling personality traits in experiments.

Core claim

Personality engineering uses AI agents to precisely parameterize, manipulate, and evaluate negotiator personality along the two dimensions of the interpersonal circumplex—warmth and dominance—providing both a rigorous methodology for testing classic negotiation theories and a practical guide for designing the personalities of AI negotiation agents.

What carries the argument

Personality engineering methodology, which parameterizes negotiator personality using the warmth and dominance dimensions of the interpersonal circumplex as a coordinate system.

If this is right

Researchers can now test canonical negotiation prescriptions under conditions where personality is precisely controlled rather than subject to human inconsistency.
AI agents can be deployed as negotiation counterparts with reproducible and tunable personality traits.
The method supplies a concrete framework for designing AI negotiators that reliably balance empathy and assertiveness.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

Large-scale simulations could systematically vary personality combinations to map their effects on multi-party negotiations.
The same parameterization might be used to create training environments in which humans practice against AI agents exhibiting targeted traits.
The approach could be extended to examine interactions between personality dimensions and external factors such as time pressure or information asymmetry.

Load-bearing premise

The interpersonal circumplex and its two core dimensions of warmth and dominance provide a suitable foundational coordinate system for parameterizing and manipulating negotiator personality in AI agents.

What would settle it

An experiment in which AI agents set to particular warmth and dominance values produce negotiation behaviors or outcomes that contradict the predictions of existing theory for those personality profiles.

read the original abstract

According to canonical negotiation theory, people's success in a negotiation depends on how well they balance competing demands--empathizing and asserting, demonstrating concern for other and concern for self, being soft on the people and hard on the problem. Yet people struggle to manage these tensions, so researchers have lacked the ability to rigorously test the field's prescriptions under controlled conditions. AI agents do not face the same limitations, and their precision, repertoire, consistency, and scalability enable a new class of experiments to contribute to negotiation theory. In this article, we introduce personality engineering: a methodology that uses AI agents to precisely parameterize, manipulate, and evaluate negotiator personality. We propose using the interpersonal circumplex--and its two core dimensions of warmth and dominance--as a foundational coordinate system for the field. This approach offers both a rigorous methodology for testing classic negotiation theories and a practical guide for designing the personalities of AI negotiation agents.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

This paper proposes using AI agents parameterized by the interpersonal circumplex to run controlled negotiation experiments, but it is still a conceptual outline without data or validation.

read the letter

This paper's core contribution is a proposed methodology called personality engineering, where AI agents are parameterized along the interpersonal circumplex dimensions of warmth and dominance to simulate negotiators for controlled experiments. It aims to test ideas from negotiation theory that are hard to isolate with human participants due to their variability. The main pitch is that AI can deliver the precision and consistency that people lack, opening up experiments on balancing empathy and assertiveness that have been difficult to run cleanly before. That framing is the one or two things worth knowing upfront. The approach is new in framing AI agents as tools for precise manipulation of personality traits in negotiation settings, drawing on the circumplex model to balance concerns for self and other. The paper does a good job explaining the limitations of traditional human-subject research and how AI's consistency and scalability could allow for better tests of prescriptions like being soft on people but hard on the problem. It engages honestly with why existing methods fall short and sketches a coordinate system that could make manipulations more reproducible. However, there are soft spots in the argument. The assumption that warmth and dominance will sufficiently capture and control the key behavioral variations seems optimistic. Factors such as BATNA, time preferences, and information asymmetry may not align cleanly with these axes, potentially confounding results unless the methodology explicitly orthogonalizes them or demonstrates their dominance through data. Since the manuscript appears to be mostly a proposal without experiments, error analysis, or validation, it's difficult to assess how well this works in practice. The stress-test concern about unmodeled variables holds up on the current evidence. This kind of work is aimed at negotiation researchers and those applying AI to social science experiments. Readers who want to explore new methods for theory testing could find the framework useful as a starting point, even if they need to fill in the implementation details themselves. I would recommend putting this through peer review. The idea has enough promise to warrant feedback from experts, particularly on how to handle the control issues and add empirical components, rather than desk rejecting it outright.

Referee Report

2 major / 0 minor

Summary. The paper proposes 'personality engineering' as a methodology that leverages AI agents to precisely parameterize, manipulate, and evaluate negotiator personality using the interpersonal circumplex (warmth and dominance dimensions). This is intended to overcome human limitations in consistency and scalability, enabling rigorous tests of canonical negotiation prescriptions such as balancing empathy with assertiveness and concern for self with concern for other.

Significance. If the circumplex parameterization can be shown to isolate personality effects while controlling for other negotiation variables, the approach would provide a valuable new experimental toolkit for negotiation theory and practical guidance for AI agent design. The conceptual framing is a strength, but the manuscript's significance is currently limited by its purely methodological nature without validation.

major comments (2)

The central claim that the interpersonal circumplex provides a suitable foundational coordinate system requires demonstration that warmth and dominance capture or control the dominant sources of behavioral variation. The skeptic's concern is valid here: variables such as BATNA, reservation values, time preferences, and information asymmetry do not map cleanly onto these axes and could interact with prompt-based instructions, undermining attribution of outcomes to personality.
No empirical validation, pilot experiments, or error analysis is presented to show that circumplex-based agent instructions produce the intended personality manipulations while holding other factors fixed. This is load-bearing for the assertion that AI agents enable a 'new class of experiments' to contribute to negotiation theory.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback on our manuscript. The comments highlight important considerations for establishing the proposed methodology's validity. We respond to each major comment below and describe the revisions we will incorporate.

read point-by-point responses

Referee: The central claim that the interpersonal circumplex provides a suitable foundational coordinate system requires demonstration that warmth and dominance capture or control the dominant sources of behavioral variation. The skeptic's concern is valid here: variables such as BATNA, reservation values, time preferences, and information asymmetry do not map cleanly onto these axes and could interact with prompt-based instructions, undermining attribution of outcomes to personality.

Authors: We agree that the suitability of the interpersonal circumplex as a coordinate system ultimately requires empirical demonstration that it can isolate personality effects amid other negotiation variables. The manuscript draws on established psychological literature to position warmth and dominance as core dimensions, but we recognize that variables like BATNA and information asymmetry may interact with personality prompts. In the revised version, we will add a new subsection in the Discussion that explicitly addresses these potential interactions and outlines experimental controls (e.g., holding BATNA, reservation values, and time preferences fixed across conditions) to support attribution to the circumplex dimensions. This will strengthen the methodological framing without overstating current evidence. revision: yes
Referee: No empirical validation, pilot experiments, or error analysis is presented to show that circumplex-based agent instructions produce the intended personality manipulations while holding other factors fixed. This is load-bearing for the assertion that AI agents enable a 'new class of experiments' to contribute to negotiation theory.

Authors: The manuscript is a conceptual and methodological proposal rather than an empirical study, so it does not include validation experiments or error analysis. We accept that this limits the immediate strength of claims about enabling rigorous tests of negotiation theory. We will revise the manuscript to include a dedicated section proposing a pilot study design, including specific protocols for verifying that circumplex-based prompts achieve intended warmth and dominance levels (e.g., via independent behavioral coding) while controlling other factors. This addition will provide a concrete roadmap for the validation needed to support the methodology's contribution to the field. revision: yes

Circularity Check

0 steps flagged

No significant circularity in methodology proposal

full rationale

The paper proposes a new methodology called personality engineering that uses AI agents to parameterize and manipulate negotiator personality via the interpersonal circumplex dimensions of warmth and dominance. This is presented as an enabling framework for testing classic negotiation theories under controlled conditions, without any equations, fitted parameters, predictions, or derivations that reduce to their own inputs by construction. The choice of the circumplex is explicitly proposed as a foundational coordinate system rather than derived from self-citations, uniqueness theorems, or smuggled ansatzes. No self-definitional loops, renamed empirical patterns, or load-bearing self-references appear in the abstract or described structure. The central claim remains a conceptual suggestion for new experiments and is self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

Review is based solely on the abstract; no specific free parameters, additional axioms, or invented entities are detailed beyond the core proposal.

axioms (1)

domain assumption The interpersonal circumplex with warmth and dominance dimensions is an appropriate foundational coordinate system for negotiator personality.
Invoked in the abstract as the basis for the personality engineering methodology.

pith-pipeline@v0.9.0 · 5688 in / 1218 out tokens · 43999 ms · 2026-05-21T06:18:01.828238+00:00 · methodology

discussion (0)

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

IndisputableMonolith/Cost/FunctionalEquation.lean washburn_uniqueness_aczel unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

We propose using the interpersonal circumplex—and its two core dimensions of warmth and dominance—as a foundational coordinate system for the field.
IndisputableMonolith/Foundation/AlexanderDuality.lean alexander_duality_circle_linking unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

AI agents do not face the same limitations, and their precision, repertoire, consistency, and scalability enable a new class of experiments

What do these tags mean?

matches: The paper's claim is directly supported by a theorem in the formal canon.
supports: The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends: The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses: The paper appears to rely on the theorem as machinery.
contradicts: The paper's claim conflicts with a theorem or certificate in the canon.
unclear: Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

4 extracted references · 4 canonical work pages

[1]

Vaccaro* and Jared R

Personality Engineering with AI Agents: A New Methodology for Negotiation Research Michelle A. Vaccaro* and Jared R. Curhan^ May 19, 2026 Abstract According to canonical negotiation theory, people’s success in a negotiation depends on how well they balance competing demands—empathizing and asserting, demonstrating concern for other and concern for self, b...

work page 2026
[2]

as our coordinate system, parameterizing AI agents along its two core dimensions—warmth and dominance—which we map onto the dual imperatives that negotiation theory has long prescribed but struggled to test (Fisher & Ury, 1981; Pruitt & Rubin, 1986; Mnookin, Peppet, & Tulumello, 2000). We describe how, unlike human negotiators, AI agents can be positioned...

work page 1981
[3]

somewhat warm and somewhat dominant

The interpersonal circumplex as a coordinate system for personality engineering. The left panel illustrates the gap between theory and practice in negotiations with humans: negotiation theory has long prescribed that effective negotiators must balance competing interpersonal demands, but human negotiators tend to perceive these as tradeoffs, making it dif...

work page doi:10.1038/s42256-025-01115-6 1995
[4]

R., Prietula, M

Weingart, L. R., Prietula, M. J., Hyder, E. B., & Genovese, C. R. (1999). Knowledge and the sequential processes of negotiation: A Markov chain analysis of response-in-kind. Journal of Experimental Social Psychology , 35 (4), 366–393. *Wiggins, J. S. (1979). A psychological taxonomy of trait-descriptive terms: The interpersonal domain. Journal of Personal...

work page 1999

[1] [1]

Vaccaro* and Jared R

Personality Engineering with AI Agents: A New Methodology for Negotiation Research Michelle A. Vaccaro* and Jared R. Curhan^ May 19, 2026 Abstract According to canonical negotiation theory, people’s success in a negotiation depends on how well they balance competing demands—empathizing and asserting, demonstrating concern for other and concern for self, b...

work page 2026

[2] [2]

as our coordinate system, parameterizing AI agents along its two core dimensions—warmth and dominance—which we map onto the dual imperatives that negotiation theory has long prescribed but struggled to test (Fisher & Ury, 1981; Pruitt & Rubin, 1986; Mnookin, Peppet, & Tulumello, 2000). We describe how, unlike human negotiators, AI agents can be positioned...

work page 1981

[3] [3]

somewhat warm and somewhat dominant

The interpersonal circumplex as a coordinate system for personality engineering. The left panel illustrates the gap between theory and practice in negotiations with humans: negotiation theory has long prescribed that effective negotiators must balance competing interpersonal demands, but human negotiators tend to perceive these as tradeoffs, making it dif...

work page doi:10.1038/s42256-025-01115-6 1995

[4] [4]

R., Prietula, M

Weingart, L. R., Prietula, M. J., Hyder, E. B., & Genovese, C. R. (1999). Knowledge and the sequential processes of negotiation: A Markov chain analysis of response-in-kind. Journal of Experimental Social Psychology , 35 (4), 366–393. *Wiggins, J. S. (1979). A psychological taxonomy of trait-descriptive terms: The interpersonal domain. Journal of Personal...

work page 1999