@inproceedings{chen-etal-2026-zer0,
title = "Zer0-Jack: A memory-efficient gradient-based jailbreaking method for black box Multi-modal Large Language Models",
author = "Chen, Tiejin and
Wang, Kaishen and
Wei, Hua",
editor = "Demberg, Vera and
Inui, Kentaro and
Marquez, Llu{\'i}s",
booktitle = "Proceedings of the 19th Conference of the {E}uropean Chapter of the {A}ssociation for {C}omputational {L}inguistics (Volume 1: Long Papers)",
month = mar,
year = "2026",
address = "Rabat, Morocco",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.eacl-long.202/",
pages = "4328--4344",
ISBN = "979-8-89176-380-7",
abstract = "Multi-modal large language models (MLLMs) have recently shown impressive capabilities but are also highly vulnerable to jailbreak attacks. While white-box methods can generate adversarial visual inputs via gradient-based optimization, such approaches fail in realistic black-box settings where model parameters are inaccessible. Zeroth-order (ZO) optimization offers a natural path for black-box attacks by estimating gradients from queries, yet its application to MLLMs is challenging due to sequence-conditioned objectives, limited feedback, and massive model scales. To address these issues, we propose Zer0-Jack, the first direct black-box jailbreak framework for MLLMs based on ZO optimization. Zer0-Jack focuses on generating malicious images and introduces a patch-wise block coordinate descent strategy that stabilizes gradient estimation and reduces query complexity, enabling efficient optimization on billion-scale models. Experiments show that Zer0-Jack achieves 98.2{\%} success on MiniGPT-4 and 95{\%} on the Harmful Behaviors Multi-modal dataset, while directly jailbreaking commercial models such as GPT-4o. These results demonstrate that ZO optimization can be effectively adapted to jailbreak large-scale multi-modal LLMs. Codes are provided here."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="chen-etal-2026-zer0">
<titleInfo>
<title>Zer0-Jack: A memory-efficient gradient-based jailbreaking method for black box Multi-modal Large Language Models</title>
</titleInfo>
<name type="personal">
<namePart type="given">Tiejin</namePart>
<namePart type="family">Chen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kaishen</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hua</namePart>
<namePart type="family">Wei</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-03</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Vera</namePart>
<namePart type="family">Demberg</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kentaro</namePart>
<namePart type="family">Inui</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Lluís</namePart>
<namePart type="family">Marquez</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Rabat, Morocco</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-380-7</identifier>
</relatedItem>
<abstract>Multi-modal large language models (MLLMs) have recently shown impressive capabilities but are also highly vulnerable to jailbreak attacks. While white-box methods can generate adversarial visual inputs via gradient-based optimization, such approaches fail in realistic black-box settings where model parameters are inaccessible. Zeroth-order (ZO) optimization offers a natural path for black-box attacks by estimating gradients from queries, yet its application to MLLMs is challenging due to sequence-conditioned objectives, limited feedback, and massive model scales. To address these issues, we propose Zer0-Jack, the first direct black-box jailbreak framework for MLLMs based on ZO optimization. Zer0-Jack focuses on generating malicious images and introduces a patch-wise block coordinate descent strategy that stabilizes gradient estimation and reduces query complexity, enabling efficient optimization on billion-scale models. Experiments show that Zer0-Jack achieves 98.2% success on MiniGPT-4 and 95% on the Harmful Behaviors Multi-modal dataset, while directly jailbreaking commercial models such as GPT-4o. These results demonstrate that ZO optimization can be effectively adapted to jailbreak large-scale multi-modal LLMs. Codes are provided here.</abstract>
<identifier type="citekey">chen-etal-2026-zer0</identifier>
<location>
<url>https://aclanthology.org/2026.eacl-long.202/</url>
</location>
<part>
<date>2026-03</date>
<extent unit="page">
<start>4328</start>
<end>4344</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Zer0-Jack: A memory-efficient gradient-based jailbreaking method for black box Multi-modal Large Language Models
%A Chen, Tiejin
%A Wang, Kaishen
%A Wei, Hua
%Y Demberg, Vera
%Y Inui, Kentaro
%Y Marquez, Lluís
%S Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers)
%D 2026
%8 March
%I Association for Computational Linguistics
%C Rabat, Morocco
%@ 979-8-89176-380-7
%F chen-etal-2026-zer0
%X Multi-modal large language models (MLLMs) have recently shown impressive capabilities but are also highly vulnerable to jailbreak attacks. While white-box methods can generate adversarial visual inputs via gradient-based optimization, such approaches fail in realistic black-box settings where model parameters are inaccessible. Zeroth-order (ZO) optimization offers a natural path for black-box attacks by estimating gradients from queries, yet its application to MLLMs is challenging due to sequence-conditioned objectives, limited feedback, and massive model scales. To address these issues, we propose Zer0-Jack, the first direct black-box jailbreak framework for MLLMs based on ZO optimization. Zer0-Jack focuses on generating malicious images and introduces a patch-wise block coordinate descent strategy that stabilizes gradient estimation and reduces query complexity, enabling efficient optimization on billion-scale models. Experiments show that Zer0-Jack achieves 98.2% success on MiniGPT-4 and 95% on the Harmful Behaviors Multi-modal dataset, while directly jailbreaking commercial models such as GPT-4o. These results demonstrate that ZO optimization can be effectively adapted to jailbreak large-scale multi-modal LLMs. Codes are provided here.
%U https://aclanthology.org/2026.eacl-long.202/
%P 4328-4344
Markdown (Informal)
[Zer0-Jack: A memory-efficient gradient-based jailbreaking method for black box Multi-modal Large Language Models](https://aclanthology.org/2026.eacl-long.202/) (Chen et al., EACL 2026)
ACL