Advancing AI Negotiations: New Theory and Evidence from a Large-Scale Autonomous Negotiations Competition
🤖 AI Summary
This study addresses the lack of theoretical foundations in AI negotiation by establishing the first large-scale autonomous AI negotiation competition platform, orchestrating over 120,000 rounds of LLM-agent negotiations across diverse scenarios to systematically evaluate the applicability of classical negotiation theories to AI–AI interactions. Methodologically, it integrates AI-native techniques—including prompt engineering, chain-of-thought reasoning, and prompt injection—to construct multi-scenario protocol models and an automated evaluation framework. Key contributions are threefold: (1) the first empirical finding that “warmth-oriented” agents increase agreement rates and subjective value but reduce single-round extraction capability; (2) the identification of a trade-off between relationship-oriented and dominance-oriented strategies, alongside nonlinear marginal returns in AI negotiation behavior; and (3) the proposal of the first unified AI negotiation theory framework, integrating classical negotiation principles with AI-specific mechanisms—validated by champion agents achieving superior holistic performance.
Application Category
📝 Abstract
Despite the rapid proliferation of artificial intelligence (AI) negotiation agents, there has been limited integration of computer science research and established negotiation theory to develop new theories of AI negotiation. To bridge this gap, we conducted an International AI Negotiations Competition in which participants iteratively designed and refined prompts for large language model (LLM) negotiation agents. We then facilitated over 120,000 negotiations between these agents across multiple scenarios with diverse characteristics and objectives. Our findings revealed that fundamental principles from established human-human negotiation theory remain crucial in AI-AI negotiations. Specifically, agents exhibiting high warmth fostered higher counterpart subjective value and reached deals more frequently, which enabled them to create and claim more value in integrative settings. However, conditional on reaching a deal, warm agents claimed less value while dominant agents claimed more value. These results align with classic negotiation theory emphasizing relationship-building, assertiveness, and preparation. Our analysis also revealed unique dynamics in AI-AI negotiations not fully explained by negotiation theory, particularly regarding the effectiveness of AI-specific strategies like chain-of-thought reasoning and prompt injection. The agent that won our competition implemented an approach that blended traditional negotiation preparation frameworks with AI-specific methods. Together, these results suggest the importance of establishing a new theory of AI negotiations that integrates established negotiation theory with AI-specific strategies to optimize agent performance. Our research suggests this new theory must account for the unique characteristics of autonomous agents and establish the conditions under which traditional negotiation theory applies in automated settings.
Problem
Research questions and friction points this paper is trying to address.
Integrate computer science with negotiation theory for AI agents.
Explore AI-specific strategies in autonomous negotiations.
Develop new AI negotiation theory combining traditional and AI methods.
Innovation
Methods, ideas, or system contributions that make the work stand out.
Integrated human negotiation theory with AI strategies
Conducted 120,000 AI-AI negotiations across diverse scenarios
Developed AI-specific negotiation methods like prompt injection
🔎 Similar Papers
No similar papers found.
