🚩 New Paper! The results of our MIT International AI Negotiations Competition are now out now in @PNASNews. We facilitated ~200,000 AI Agent Negotiations across multiple diverse scenarios and objectives in which participants iteratively designed and refined AI negotiation agents. Our findings revealed that 🚩 Fundamental principles from established human-human negotiation theory remain crucial in AI-AI negotiations. Specifically, 🚩 AI 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 (who says AI agents are only rational optimizers?). 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. 🚩 Critically, we are releasing all the prompts and all the transcripts of all the negotiations publicly to support replication and future research. Kudos to first author @MichelleVaccaro and the rest of the team including Jared Curhan @caosunova @harangju. Major thanks to cosponsors the @mit_ide and @iDecisionGames! We also thank @OpenAI for model access, and @MITSloanExecEd and the MIT Office of Teaching and Learning for institutional support. We are grateful to all participants in the MIT AI Negotiations Competition for their engagement and creativity. Their innovative approaches significantly contributed to our understanding of AI negotiation dynamics. pnas.org/doi/10.1073/pnas.25… As always, thoughts and comments highly encouraged!