![[2602.12089] Choose Your Agent: Tradeoffs in Adopting AI Advisors, Coaches, and Delegates in Multi-Party Negotiation](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.12089] Choose Your Agent: Tradeoffs in Adopting AI Advisors, Coaches, and Delegates in Multi-Party Negotiation
Summary
This study explores the tradeoffs in using AI agents in multi-party negotiations, revealing a preference-performance misalignment among users in different roles.
Why It Matters
As AI becomes integral in negotiation settings, understanding how different AI modalities affect user outcomes is crucial. This research highlights the complexities of user preferences versus actual performance, informing the design of more effective AI systems that enhance group welfare.
Key Takeaways
- Participants preferred AI Advisors but achieved better outcomes using AI Delegates.
- Delegation by AI agents can create positive externalities for non-users.
- The study identifies a gap between perceived agent capabilities and actual welfare gains.
- Effective AI design must consider user perceptions and adoption barriers.
- Autonomous agents can enhance negotiation outcomes if properly integrated.
Computer Science > Computer Science and Game Theory arXiv:2602.12089 (cs) [Submitted on 12 Feb 2026 (v1), last revised 13 Feb 2026 (this version, v2)] Title:Choose Your Agent: Tradeoffs in Adopting AI Advisors, Coaches, and Delegates in Multi-Party Negotiation Authors:Kehang Zhu, Nithum Thain, Vivian Tsai, James Wexler, Crystal Qian View a PDF of the paper titled Choose Your Agent: Tradeoffs in Adopting AI Advisors, Coaches, and Delegates in Multi-Party Negotiation, by Kehang Zhu and 4 other authors View PDF HTML (experimental) Abstract:As AI usage becomes more prevalent in social contexts, understanding agent-user interaction is critical to designing systems that improve both individual and group outcomes. We present an online behavioral experiment (N = 243) in which participants play three multi-turn bargaining games in groups of three. Each game, presented in randomized order, grants access to a single LLM assistance modality: proactive recommendations from an Advisor, reactive feedback from a Coach, or autonomous execution by a Delegate; all modalities are powered by an underlying LLM that achieves superhuman performance in an all-agent environment. On each turn, participants privately decide whether to act manually or use the AI modality available in that game. Despite preferring the Advisor modality, participants achieve the highest mean individual gains with the Delegate, demonstrating a preference-performance misalignment. Moreover, delegation generates positive ex...
