Skip to main content
IACM 2024

IACM 2024 Abstract Book »

Smooth-Talking Bots: AI Negotiators Make Better Impressions


Abstract: Employing an online experiment and chatbot built using GPT-4, we conducted 203 trials of a distributive bargaining simulation (N= 231) across a variety of conditions that systematically manipulated the AI chatbot's warmth and dominance, assessing how objective and subjective negotiation outcomes differed between human vs. human and human vs. chatbot negotiation conditions. Results indicated no significant differences in objective outcomes across conditions. However, subjective value reported by the human was higher when facing chatbots than when facing other humans. The study highlights the impact of AI on negotiation dynamics, offering insights into human perceptions and behaviors when interacting with AI counterparts.

Keywords: AI, Negotiations, Experiments, Chatbot, Subjective Value

David Fang, Stanford Graduate School of Business (United States)
Email: davfang@stanford.edu

Mohammed Alsobay, MIT Sloan (United States)
Email: mosobay@mit.edu

Abdullah Almaatouq, MIT Sloan (United States)
Email: amaatouq@mit.edu

Jared Curhan, MIT Sloan (United States)
Email: curhan@mit.edu

 


Powered by OpenConf®
Copyright ©2002-2023 Zakon Group LLC