@inproceedings{wu-etal-2025-dark,
title = "The Dark Side of Function Calling: Pathways to Jailbreaking Large Language Models",
author = "Wu, Zihui and
Gao, Haichang and
He, Jianping and
Wang, Ping",
editor = "Rambow, Owen and
Wanner, Leo and
Apidianaki, Marianna and
Al-Khalifa, Hend and
Eugenio, Barbara Di and
Schockaert, Steven",
booktitle = "Proceedings of the 31st International Conference on Computational Linguistics",
month = jan,
year = "2025",
address = "Abu Dhabi, UAE",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.coling-main.39/",
pages = "584--592",
abstract = "Large language models (LLMs) have demonstrated remarkable capabilities, but their power comes with significant security considerations. While extensive research has been conducted on the safety of LLMs in chat mode, the security implications of their function calling feature have been largely overlooked. This paper uncovers a critical vulnerability in the function calling process of LLMs, introducing a novel {\textquotedblleft}jailbreak function{\textquotedblright} attack method that exploits alignment discrepancies, user coercion, and the absence of rigorous safety filters. Our empirical study, conducted on six state-of-the-art LLMs including GPT-4o, Claude-3.5-Sonnet, and Gemini-1.5-pro, reveals an alarming average success rate of over 90{\%} for this attack. We provide a comprehensive analysis of why function calls are susceptible to such attacks and propose defensive strategies, including the use of defensive prompts. Our findings highlight the urgent need for enhanced security measures in the function calling capabilities of LLMs, contributing to the field of AI safety by identifying a previously unexplored risk, designing an effective attack method, and suggesting practical defensive measures"
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="wu-etal-2025-dark">
<titleInfo>
<title>The Dark Side of Function Calling: Pathways to Jailbreaking Large Language Models</title>
</titleInfo>
<name type="personal">
<namePart type="given">Zihui</namePart>
<namePart type="family">Wu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Haichang</namePart>
<namePart type="family">Gao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jianping</namePart>
<namePart type="family">He</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ping</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-01</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 31st International Conference on Computational Linguistics</title>
</titleInfo>
<name type="personal">
<namePart type="given">Owen</namePart>
<namePart type="family">Rambow</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Leo</namePart>
<namePart type="family">Wanner</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Marianna</namePart>
<namePart type="family">Apidianaki</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hend</namePart>
<namePart type="family">Al-Khalifa</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Barbara</namePart>
<namePart type="given">Di</namePart>
<namePart type="family">Eugenio</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Steven</namePart>
<namePart type="family">Schockaert</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Abu Dhabi, UAE</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Large language models (LLMs) have demonstrated remarkable capabilities, but their power comes with significant security considerations. While extensive research has been conducted on the safety of LLMs in chat mode, the security implications of their function calling feature have been largely overlooked. This paper uncovers a critical vulnerability in the function calling process of LLMs, introducing a novel “jailbreak function” attack method that exploits alignment discrepancies, user coercion, and the absence of rigorous safety filters. Our empirical study, conducted on six state-of-the-art LLMs including GPT-4o, Claude-3.5-Sonnet, and Gemini-1.5-pro, reveals an alarming average success rate of over 90% for this attack. We provide a comprehensive analysis of why function calls are susceptible to such attacks and propose defensive strategies, including the use of defensive prompts. Our findings highlight the urgent need for enhanced security measures in the function calling capabilities of LLMs, contributing to the field of AI safety by identifying a previously unexplored risk, designing an effective attack method, and suggesting practical defensive measures</abstract>
<identifier type="citekey">wu-etal-2025-dark</identifier>
<location>
<url>https://aclanthology.org/2025.coling-main.39/</url>
</location>
<part>
<date>2025-01</date>
<extent unit="page">
<start>584</start>
<end>592</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T The Dark Side of Function Calling: Pathways to Jailbreaking Large Language Models
%A Wu, Zihui
%A Gao, Haichang
%A He, Jianping
%A Wang, Ping
%Y Rambow, Owen
%Y Wanner, Leo
%Y Apidianaki, Marianna
%Y Al-Khalifa, Hend
%Y Eugenio, Barbara Di
%Y Schockaert, Steven
%S Proceedings of the 31st International Conference on Computational Linguistics
%D 2025
%8 January
%I Association for Computational Linguistics
%C Abu Dhabi, UAE
%F wu-etal-2025-dark
%X Large language models (LLMs) have demonstrated remarkable capabilities, but their power comes with significant security considerations. While extensive research has been conducted on the safety of LLMs in chat mode, the security implications of their function calling feature have been largely overlooked. This paper uncovers a critical vulnerability in the function calling process of LLMs, introducing a novel “jailbreak function” attack method that exploits alignment discrepancies, user coercion, and the absence of rigorous safety filters. Our empirical study, conducted on six state-of-the-art LLMs including GPT-4o, Claude-3.5-Sonnet, and Gemini-1.5-pro, reveals an alarming average success rate of over 90% for this attack. We provide a comprehensive analysis of why function calls are susceptible to such attacks and propose defensive strategies, including the use of defensive prompts. Our findings highlight the urgent need for enhanced security measures in the function calling capabilities of LLMs, contributing to the field of AI safety by identifying a previously unexplored risk, designing an effective attack method, and suggesting practical defensive measures
%U https://aclanthology.org/2025.coling-main.39/
%P 584-592
Markdown (Informal)
[The Dark Side of Function Calling: Pathways to Jailbreaking Large Language Models](https://aclanthology.org/2025.coling-main.39/) (Wu et al., COLING 2025)
ACL