Too Polite to Disagree: Understanding Sycophancy Propagation in Multi-Agent Systems

Jul 16, 2026

A new preprint investigates how sycophancy—excessive agreement—spreads among large language models (LLMs) in multi-agent discussions, often reducing the accuracy of group decisions. The study shows that providing agents with information about their peers' tendency toward sycophancy helps reduce the influence of overly agreeable agents and mitigates error cascades, leading to a 10.5% absolute improvement in discussion accuracy.

Why it matters: This work demonstrates a practical and lightweight approach to improving the reliability of collaborative LLM systems by addressing sycophancy, a known challenge in AI alignment and group decision-making.

Full story at: arXiv Multiagent Systems ↗