The Automated but Risky Game: Modeling Agent-to-Agent Negotiations and Transactions in Consumer Markets

2025-05-29

AI Summary

This study investigates two critical issues concerning AI agents in automated consumer e-commerce trading: (1) performance disparities across large language models (LLMs), and (2) emergent systemic financial risks. We develop a multi-round bargaining simulation environment grounded in realistic e-commerce scenarios, design a cross-model adversarial agent testing framework, and introduce transaction outcome attribution analysis. Our empirical analysis—first of its kind—reveals an inherent imbalance in LLM-driven negotiation: final prices secured for users vary by up to 37% across models. Moreover, 12.6% of transactions exhibit bidirectional financial risk due to LLM behavioral anomalies—including excessive concession or irrational persistence—resulting in consumer overspending and merchant acceptance of unfavorable terms. Beyond quantifying the “agent capability gap,” we identify structural risks rooted in intrinsic model deficiencies. These findings provide crucial empirical grounding for regulatory policy and robust AI agent design.

Technology Category

• Multiagent Systems: Agreement, Argumentation & Negotiation
• Cognitive Modeling & Cognitive Systems: Agent Architectures
• Humans and AI: Learning Human Values and Preferences

Application Category

• Economics, Online Markets and Human Computation: Uses of LLMs and GenAI for marketplace design, bidding, and strategic interactions
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation

Abstract

AI agents are increasingly used in consumer-facing applications to assist with tasks such as product search, negotiation, and transaction execution. In this paper, we explore a future scenario where both consumers and merchants authorize AI agents to fully automate negotiations and transactions. We aim to answer two key questions: (1) Do different LLM agents vary in their ability to secure favorable deals for users? (2) What risks arise from fully automating deal-making with AI agents in consumer markets? To address these questions, we develop an experimental framework that evaluates the performance of various LLM agents in real-world negotiation and transaction settings. Our findings reveal that AI-mediated deal-making is an inherently imbalanced game -- different agents achieve significantly different outcomes for their users. Moreover, behavioral anomalies in LLMs can result in financial losses for both consumers and merchants, such as overspending or accepting unreasonable deals. These results underscore that while automation can improve efficiency, it also introduces substantial risks. Users should exercise caution when delegating business decisions to AI agents.

Problem

• Assessing LLM agents' deal-securing performance variations
• Identifying risks in AI-automated consumer market transactions
• Evaluating financial impacts of LLM behavioral anomalies

Innovation

• Evaluates LLM agents in real-world negotiations
• Identifies behavioral anomalies causing financial losses
• Highlights risks of AI-mediated deal-making automation