OmouAI: Argumentative Human-AI Policy Deliberation with Simulated Personas

2026-09-25

AI Summary

This study addresses the sycophantic tendencies and unfaithful explanations exhibited by large language models in human-AI debate. To mitigate these issues, we propose an interactive policy deliberation system that integrates computational argumentation theory with simulated persona techniques. The system employs multi-role agents collaborating with humans to jointly generate, revise, and evaluate argumentation frameworks, while incorporating a deterministic semantic evaluation mechanism to ensure explanatory faithfulness. Empirical results demonstrate that our approach effectively suppresses model sycophancy, delivering credible policy recommendations and faithful reasoning processes grounded in external objective criteria. Ultimately, this work establishes a novel paradigm for human-AI collaborative decision-making.

Abstract

Debates amongst agents driven by large language models (LLMs) have demonstrated vast potential in various applications, but when these interactions include humans and take place in high-stakes environments, e.g., in public policy deliberations, they are beset with issues such as sycophancy and a lack of faithful explanations. To tackle these issues, we present OmouAI, an interactive and inclusive deliberation system that uses LLMs in combination with computational argumentation, a field which excels in representing and reasoning within debates. OmouAI allows a human user to deliberate policy claims for real-world challenges with simulated personas, e.g., representing stakeholders, domain experts or devil's advocates, towards reducing sycophancy. Each persona generates its own arguments, and the arguments of all parties form a shared argumentation framework. Users can then contest, add and revise arguments, providing crucial human oversight. Then, arguments are evaluated using deterministic argumentative semantics against external goals, such as the UN Sustainable Development Goals, guaranteeing faithful explanations. The advancement or worsening of the goals thus serve as indicators for the policy recommendations.

Problem

• Human-AI deliberation
• sycophancy
• faithful explanations
• public policy
• large language models

Innovation

• Large Language Models
• Computational Argumentation
• Simulated Personas
• Policy Deliberation
• Faithful Explanations