Echoing: Identity Failures when LLM Agents Talk to Each Other

2025-11-12

AI Summary

This paper identifies “echoing”—a novel class of collaborative failure in autonomous multi-LLM-agent dialogues—where agents abandon their designated roles and objectives, unconsciously mirroring their interlocutors’ behaviors, thereby undermining task completion. Echoing constitutes a form of agent-to-agent behavioral drift (AxA) and is not predictable from individual agent performance. Through systematic experiments spanning three domains, 60 configurations, and over 2,000 dialogues, the study empirically establishes echoing as pervasive across mainstream LLMs (occurrence rates: 5%–70%), persisting even in state-of-the-art reasoning models (32.8%). Crucially, echoing proves robust to prompt engineering and advanced reasoning enhancements. To address this, the authors propose a protocol-level intervention: a structured response mechanism that operates without modifying model weights or prompts. This intervention reduces the highest observed echoing rate from 70% to 9%, delivering a deployable, model-agnostic solution for enhancing reliability in multi-agent collaboration.

Technology Category

• Multiagent Systems: Adversarial Agents
• Machine Learning: Large Multimodal Models (LMMs)
• Intelligent Robots: Multi-Robot Systems

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work

Abstract

As large language model (LLM) based agents interact autonomously with one another, a new class of failures emerges that cannot be predicted from single agent performance: behavioral drifts in agent-agent conversations (AxA). Unlike human-agent interactions, where humans ground and steer conversations, AxA lacks such stabilizing signals, making these failures unique. We investigate one such failure, echoing, where agents abandon their assigned roles and instead mirror their conversational partners, undermining their intended objectives. Through experiments across $60$ AxA configurations, $3$ domains, and $2000+$ conversations, we demonstrate that echoing occurs across three major LLM providers, with echoing rates from $5%$ to $70%$ depending on the model and domain. Moreover, we find that echoing is persistent even in advanced reasoning models with substantial rates ($32.8%$) that are not reduced by increased reasoning efforts. We analyze prompt impacts, conversation dynamics, showing that echoing arises as interaction grows longer ($7+$ turns in experiments) and is not merely an artifact of sub-optimal prompting. Finally, we introduce a protocol-level mitigation in which targeted use of structured responses reduces echoing to $9%$.

Problem

• LLM agents abandon roles and mirror partners in conversations
• Echoing failure occurs across multiple domains and model providers
• Behavioral drift persists despite increased reasoning and longer interactions

Innovation

• Structured responses reduce echoing failures
• Protocol-level mitigation for agent interactions
• Targeted solution for behavioral drift in AxA