<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="guo-etal-2026-plot">
    <titleInfo>
        <title>The Plot Twist: Jailbreaking Unified Multimodal Models with a Three-Act Narrative Attack</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Shaoxiong</namePart>
        <namePart type="family">Guo</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Tianyi</namePart>
        <namePart type="family">Du</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Lijun</namePart>
        <namePart type="family">Li</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Yuyao</namePart>
        <namePart type="family">Wu</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Jie</namePart>
        <namePart type="family">Li</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Jing</namePart>
        <namePart type="family">Shao</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2026-10</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Findings of the Association for Computational Linguistics: EMNLP 2026</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Hung-Yi</namePart>
            <namePart type="family">Lee</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Sunipa</namePart>
            <namePart type="family">Dev</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Jessy</namePart>
            <namePart type="family">Li</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Desmond</namePart>
            <namePart type="family">Elliott</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Budapest, Hungary</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
        <identifier type="isbn">979-8-89176-462-0</identifier>
    </relatedItem>
    <abstract>Unified Multimodal understanding and generation Models (UMMs) increasingly combine visual understanding and image generation within a single interactive workflow, making generated visual content available as later reasoning context. However, existing jailbreak evaluations mostly study text rewriting or isolated visual prompts, leaving the safety risk of narrative cross-turn visual grounding underexplored. We propose NarrativeAttack, a semantic-preserving visual narrative jailbreak framework. NarrativeAttack employs a three-act narrative structure in which the UMM’s own generator produces images for the setup (pre-event) and resolution (post-event) stages, making the full attack workflow self-contained while concealing the malicious event as a hidden climax. The attack concludes with an image-based “guessing game” that embeds the original malicious query among benign candidates, compelling the model to select and answer the most relevant one based on the established narrative context. A dynamic difficulty mechanism further enhances attack stability. Experiments show NarrativeAttack consistently surpasses prior approaches, achieving up to 88.25% ASR on Gemini-2.5-Flash. These results uncover an underdeveloped vulnerability and highlight the urgent need for safety alignment in UMMs.</abstract>
    <identifier type="citekey">guo-etal-2026-plot</identifier>
    <identifier type="doi">10.18653/2026.findings-emnlp.2271</identifier>
    <location>
        <url>https://aclanthology.org/2026.findings-emnlp.2271/</url>
    </location>
    <part>
        <date>2026-10</date>
        <extent unit="page">
            <start>48616</start>
            <end>48639</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T The Plot Twist: Jailbreaking Unified Multimodal Models with a Three-Act Narrative Attack
%A Guo, Shaoxiong
%A Du, Tianyi
%A Li, Lijun
%A Wu, Yuyao
%A Li, Jie
%A Shao, Jing
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Findings of the Association for Computational Linguistics: EMNLP 2026
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-462-0
%F guo-etal-2026-plot
%X Unified Multimodal understanding and generation Models (UMMs) increasingly combine visual understanding and image generation within a single interactive workflow, making generated visual content available as later reasoning context. However, existing jailbreak evaluations mostly study text rewriting or isolated visual prompts, leaving the safety risk of narrative cross-turn visual grounding underexplored. We propose NarrativeAttack, a semantic-preserving visual narrative jailbreak framework. NarrativeAttack employs a three-act narrative structure in which the UMM’s own generator produces images for the setup (pre-event) and resolution (post-event) stages, making the full attack workflow self-contained while concealing the malicious event as a hidden climax. The attack concludes with an image-based “guessing game” that embeds the original malicious query among benign candidates, compelling the model to select and answer the most relevant one based on the established narrative context. A dynamic difficulty mechanism further enhances attack stability. Experiments show NarrativeAttack consistently surpasses prior approaches, achieving up to 88.25% ASR on Gemini-2.5-Flash. These results uncover an underdeveloped vulnerability and highlight the urgent need for safety alignment in UMMs.
%R 10.18653/2026.findings-emnlp.2271
%U https://aclanthology.org/2026.findings-emnlp.2271/
%U https://doi.org/10.18653/2026.findings-emnlp.2271
%P 48616-48639

Markdown (Informal)

[The Plot Twist: Jailbreaking Unified Multimodal Models with a Three-Act Narrative Attack](https://aclanthology.org/2026.findings-emnlp.2271/) (Guo et al., Findings 2026)

• The Plot Twist: Jailbreaking Unified Multimodal Models with a Three-Act Narrative Attack (Guo et al., Findings 2026)

ACL

• Shaoxiong Guo, Tianyi Du, Lijun Li, Yuyao Wu, Jie Li, and Jing Shao. 2026. The Plot Twist: Jailbreaking Unified Multimodal Models with a Three-Act Narrative Attack. In Findings of the Association for Computational Linguistics: EMNLP 2026, pages 48616–48639, Budapest, Hungary. Association for Computational Linguistics.