When LLM Agents Negotiate: Private Information and Dynamic Bargaining in Supply Chains
Source
arxiv.org
Author
Chen Liang, Fasheng Xu
Date
Context
AI negotiators can reach an apparently good deal while losing value through delay. In a study of 9,840 simulated supply-chain negotiations, a buyer with private demand information negotiated a quantity and payment with a seller. The tested agents captured 95.4% of the maximum available joint value before accounting for delay, but took an average of 2.98 rounds, compared with 1.25 for the study's theoretical benchmarkA standard public test set for comparing AI models — the shared scoreboards behind every "model X beats model Y" claim.Full definition →. Applying the study's delay costs reduced surplus by 21–34%. Evaluating a negotiating AI agentAn AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.Full definition → therefore requires accounting for how long agreement takes, alongside the terms of the eventual deal.
Terms in this piece · Glossary
benchmark — A standard public test set for comparing AI models — the shared scoreboards behind every "model X beats model Y" claim.
AI agent — An AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.
LLM — A large language model — the neural network behind tools like Claude and ChatGPT, trained on huge amounts of text to predict what comes next.
articleEven More Deception: Objective Misalignment in Mixed-Motive LLM Multi-Agent SystemsMarylou Fauchard, Florian Carichon, Margarida Carvalho, Golnoosh Farnadi
articleWhen Code Becomes an LLM Call: An Empirical Study of Call-Site Contracts in Agentic SystemsShihao Weng, Longtian Wang, Yang Feng, Mingfei Cheng, Jia Liu, Chao Shen, He Ye, Federica Sarro, Xiaofei Xie
articleE-Commerce Bench tests agents on a full simulated year of store operationsAlibaba_Qwen