@inproceedings{ceccatelli-etal-2026-speechjbb,
title = "{S}peech{JBB}: Probing Safety Alignment and Comprehension in Large Audio Language Models under Code-Switched Speech",
author = "Ceccatelli, Virginia and
Jeon, Yejin and
Adelani, David Ifeoluwa",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {EMNLP} 2026",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-emnlp.449/",
pages = "9432--9447",
ISBN = "979-8-89176-462-0",
abstract = "Large audio language models (LALMs) are increasingly deployed in real-world applications, yet their safety alignment is still primarily evaluated on monolingual, text-based harmful prompts. This leaves their generalizability under multilingual and spoken settings, particularly code-switched speech, largely underexplored. To address this gap, we introduce SpeechJBB, an audio jailbreak dataset for benchmarking state-of-the-art LALMs across five European languages: English, French, German, Italian, and Spanish, as well as code-switched variants combining pairs of these languages. The extent of safety weaknesses is further probed by introducing an augmented setting where phonologically plausible pseudo-words are inserted around safety-critical terms to simulate localized obfuscation. Across models, code-switched harmful audio yields substantially high jailbreak success rates (JSR), with non-English monolingual and non-English code-switched pairs exhibiting the highest attack success. Pseudo-word insertion monotonically reduces refusal as insertion density increases, even though models rarely attribute harmful meaning to the inserted tokens. Comprehension benchmarks show that these failures are not reducible to multilingual misunderstanding, as several models with strong ASR, spoken language understanding, and spoken reasoning performance are among the most vulnerable."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="ceccatelli-etal-2026-speechjbb">
<titleInfo>
<title>SpeechJBB: Probing Safety Alignment and Comprehension in Large Audio Language Models under Code-Switched Speech</title>
</titleInfo>
<name type="personal">
<namePart type="given">Virginia</namePart>
<namePart type="family">Ceccatelli</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yejin</namePart>
<namePart type="family">Jeon</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">David</namePart>
<namePart type="given">Ifeoluwa</namePart>
<namePart type="family">Adelani</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>Large audio language models (LALMs) are increasingly deployed in real-world applications, yet their safety alignment is still primarily evaluated on monolingual, text-based harmful prompts. This leaves their generalizability under multilingual and spoken settings, particularly code-switched speech, largely underexplored. To address this gap, we introduce SpeechJBB, an audio jailbreak dataset for benchmarking state-of-the-art LALMs across five European languages: English, French, German, Italian, and Spanish, as well as code-switched variants combining pairs of these languages. The extent of safety weaknesses is further probed by introducing an augmented setting where phonologically plausible pseudo-words are inserted around safety-critical terms to simulate localized obfuscation. Across models, code-switched harmful audio yields substantially high jailbreak success rates (JSR), with non-English monolingual and non-English code-switched pairs exhibiting the highest attack success. Pseudo-word insertion monotonically reduces refusal as insertion density increases, even though models rarely attribute harmful meaning to the inserted tokens. Comprehension benchmarks show that these failures are not reducible to multilingual misunderstanding, as several models with strong ASR, spoken language understanding, and spoken reasoning performance are among the most vulnerable.</abstract>
<identifier type="citekey">ceccatelli-etal-2026-speechjbb</identifier>
<location>
<url>https://aclanthology.org/2026.findings-emnlp.449/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>9432</start>
<end>9447</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T SpeechJBB: Probing Safety Alignment and Comprehension in Large Audio Language Models under Code-Switched Speech
%A Ceccatelli, Virginia
%A Jeon, Yejin
%A Adelani, David Ifeoluwa
%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 ceccatelli-etal-2026-speechjbb
%X Large audio language models (LALMs) are increasingly deployed in real-world applications, yet their safety alignment is still primarily evaluated on monolingual, text-based harmful prompts. This leaves their generalizability under multilingual and spoken settings, particularly code-switched speech, largely underexplored. To address this gap, we introduce SpeechJBB, an audio jailbreak dataset for benchmarking state-of-the-art LALMs across five European languages: English, French, German, Italian, and Spanish, as well as code-switched variants combining pairs of these languages. The extent of safety weaknesses is further probed by introducing an augmented setting where phonologically plausible pseudo-words are inserted around safety-critical terms to simulate localized obfuscation. Across models, code-switched harmful audio yields substantially high jailbreak success rates (JSR), with non-English monolingual and non-English code-switched pairs exhibiting the highest attack success. Pseudo-word insertion monotonically reduces refusal as insertion density increases, even though models rarely attribute harmful meaning to the inserted tokens. Comprehension benchmarks show that these failures are not reducible to multilingual misunderstanding, as several models with strong ASR, spoken language understanding, and spoken reasoning performance are among the most vulnerable.
%U https://aclanthology.org/2026.findings-emnlp.449/
%P 9432-9447
Markdown (Informal)
[SpeechJBB: Probing Safety Alignment and Comprehension in Large Audio Language Models under Code-Switched Speech](https://aclanthology.org/2026.findings-emnlp.449/) (Ceccatelli et al., Findings 2026)
ACL