@inproceedings{zhang-etal-2025-allies-turn,
title = "When Allies Turn Foes: Exploring Group Characteristics of {LLM}-Based Multi-Agent Collaborative Systems Under Adversarial Attacks",
author = "Zhang, Jiahao and
Kan, Baoshuo and
Gong, Tao and
Wang, Fu Lee and
Hao, Tianyong",
editor = "Christodoulopoulos, Christos and
Chakraborty, Tanmoy and
Rose, Carolyn and
Peng, Violet",
booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2025",
month = nov,
year = "2025",
address = "Suzhou, China",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.findings-emnlp.333/",
doi = "10.18653/v1/2025.findings-emnlp.333",
pages = "6275--6300",
ISBN = "979-8-89176-335-7",
abstract = "This paper investigates the group characteristics in multi-agent collaborative systems under adversarial attacks. Adversarial agents are tasked with generating counterfactual answers to a given collaborative problem, while collaborative agents normally interact with other agents to solve the given problem. To simulate real-world collaboration scenarios as closely as possible, we evaluate the collaborative system in three different collaboration scenarios and design three different communication strategies and different group structures. Furthermore, we explored several methods to mitigate adversarial attacks, all of which have been proven effective through our experiments. To quantify the robustness of collaborative systems against such attacks, a novel metric, System Defense Index (SDI), is introduced. Finally, we conducted an in-depth analysis from the perspective of group dynamics on how adversarial agents affect multi-agent collaborative systems, which reveals similarities between the agent collaboration process and human collaboration process. The code will be made available after publication."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="zhang-etal-2025-allies-turn">
<titleInfo>
<title>When Allies Turn Foes: Exploring Group Characteristics of LLM-Based Multi-Agent Collaborative Systems Under Adversarial Attacks</title>
</titleInfo>
<name type="personal">
<namePart type="given">Jiahao</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Baoshuo</namePart>
<namePart type="family">Kan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tao</namePart>
<namePart type="family">Gong</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Fu</namePart>
<namePart type="given">Lee</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tianyong</namePart>
<namePart type="family">Hao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-11</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: EMNLP 2025</title>
</titleInfo>
<name type="personal">
<namePart type="given">Christos</namePart>
<namePart type="family">Christodoulopoulos</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tanmoy</namePart>
<namePart type="family">Chakraborty</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Carolyn</namePart>
<namePart type="family">Rose</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Violet</namePart>
<namePart type="family">Peng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Suzhou, China</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-335-7</identifier>
</relatedItem>
<abstract>This paper investigates the group characteristics in multi-agent collaborative systems under adversarial attacks. Adversarial agents are tasked with generating counterfactual answers to a given collaborative problem, while collaborative agents normally interact with other agents to solve the given problem. To simulate real-world collaboration scenarios as closely as possible, we evaluate the collaborative system in three different collaboration scenarios and design three different communication strategies and different group structures. Furthermore, we explored several methods to mitigate adversarial attacks, all of which have been proven effective through our experiments. To quantify the robustness of collaborative systems against such attacks, a novel metric, System Defense Index (SDI), is introduced. Finally, we conducted an in-depth analysis from the perspective of group dynamics on how adversarial agents affect multi-agent collaborative systems, which reveals similarities between the agent collaboration process and human collaboration process. The code will be made available after publication.</abstract>
<identifier type="citekey">zhang-etal-2025-allies-turn</identifier>
<identifier type="doi">10.18653/v1/2025.findings-emnlp.333</identifier>
<location>
<url>https://aclanthology.org/2025.findings-emnlp.333/</url>
</location>
<part>
<date>2025-11</date>
<extent unit="page">
<start>6275</start>
<end>6300</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T When Allies Turn Foes: Exploring Group Characteristics of LLM-Based Multi-Agent Collaborative Systems Under Adversarial Attacks
%A Zhang, Jiahao
%A Kan, Baoshuo
%A Gong, Tao
%A Wang, Fu Lee
%A Hao, Tianyong
%Y Christodoulopoulos, Christos
%Y Chakraborty, Tanmoy
%Y Rose, Carolyn
%Y Peng, Violet
%S Findings of the Association for Computational Linguistics: EMNLP 2025
%D 2025
%8 November
%I Association for Computational Linguistics
%C Suzhou, China
%@ 979-8-89176-335-7
%F zhang-etal-2025-allies-turn
%X This paper investigates the group characteristics in multi-agent collaborative systems under adversarial attacks. Adversarial agents are tasked with generating counterfactual answers to a given collaborative problem, while collaborative agents normally interact with other agents to solve the given problem. To simulate real-world collaboration scenarios as closely as possible, we evaluate the collaborative system in three different collaboration scenarios and design three different communication strategies and different group structures. Furthermore, we explored several methods to mitigate adversarial attacks, all of which have been proven effective through our experiments. To quantify the robustness of collaborative systems against such attacks, a novel metric, System Defense Index (SDI), is introduced. Finally, we conducted an in-depth analysis from the perspective of group dynamics on how adversarial agents affect multi-agent collaborative systems, which reveals similarities between the agent collaboration process and human collaboration process. The code will be made available after publication.
%R 10.18653/v1/2025.findings-emnlp.333
%U https://aclanthology.org/2025.findings-emnlp.333/
%U https://doi.org/10.18653/v1/2025.findings-emnlp.333
%P 6275-6300
Markdown (Informal)
[When Allies Turn Foes: Exploring Group Characteristics of LLM-Based Multi-Agent Collaborative Systems Under Adversarial Attacks](https://aclanthology.org/2025.findings-emnlp.333/) (Zhang et al., Findings 2025)
ACL