@inproceedings{saidi-khalid-2026-l3ia,
title = "{L}3{IA} {M}orocco at {A}r{G}uard 2026: Intent over Words - Uncovering Implicit Harm in {A}rabic {LLM} Prompts",
author = "Saidi, Wiam and
Khalid, Satori",
editor = "Abdul-Mageed, Muhammad and
Elsayed, Tamer and
Bouamor, Houda and
El-Beltagy, Samhaa and
Al-Khalifa, Hend and
Abdelali, Ahmed and
Elmadany, AbdelRahim and
El Mekki, Abdellah and
Darwish, Kareem and
Alkhereyf, Sakhar and
Luqman, Hamzah and
Khalifa, Salam and
Magdy, Samar and
Alyafeai, Zaid and
Hamed, Injy and
Zaghouani, Wajdi and
Zitouni, Imed",
booktitle = "Proceedings of The Fourth {A}rabic Natural Language Processing Conference: Shared Tasks",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.arabicnlp-sharedtasks.42/",
pages = "338--346",
ISBN = "979-8-89176-497-2",
abstract = "We present the L3IA Morocco team{'}s system for shared Task B, designed for large language models and focused on detecting harmful prompts in Arabic. In this context, we define two subtasks: B1 and B2, which refer, respectively, to the binary safe/unsafe classification and the classification of harmfulness into 7 categories. Our system combines three fine-tuned Arabic encoders: AraBERTv2, CAMeLBERTmix, and MARBERT, using soft voting that takes into account the predominant classes. It achieves a macro-F1 score of 0.946 for B1 and 0.902 for B2 during the development phase. In contrast, performance drops to 0.639 for B1 and 0.758 for B2 in the blind test. We demonstrate that this generalization divergence is due to a ``unsafe'' overprediction bias resulting from an unbalanced training corpus (80{\%} unsafe) and not due to alignment of identifiers or dialects. Finally, we observed a negative result: in zeroshot and few-shot, AceGPT-7B, a generative Arabic LLM, performed poorly compared to fine-tuned encoders (0.38-0.42 on Darija)."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="saidi-khalid-2026-l3ia">
<titleInfo>
<title>L3IA Morocco at ArGuard 2026: Intent over Words - Uncovering Implicit Harm in Arabic LLM Prompts</title>
</titleInfo>
<name type="personal">
<namePart type="given">Wiam</namePart>
<namePart type="family">Saidi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Satori</namePart>
<namePart type="family">Khalid</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>Proceedings of The Fourth Arabic Natural Language Processing Conference: Shared Tasks</title>
</titleInfo>
<name type="personal">
<namePart type="given">Muhammad</namePart>
<namePart type="family">Abdul-Mageed</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tamer</namePart>
<namePart type="family">Elsayed</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Houda</namePart>
<namePart type="family">Bouamor</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Samhaa</namePart>
<namePart type="family">El-Beltagy</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">Ahmed</namePart>
<namePart type="family">Abdelali</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">AbdelRahim</namePart>
<namePart type="family">Elmadany</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Abdellah</namePart>
<namePart type="family">El Mekki</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kareem</namePart>
<namePart type="family">Darwish</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sakhar</namePart>
<namePart type="family">Alkhereyf</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hamzah</namePart>
<namePart type="family">Luqman</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Salam</namePart>
<namePart type="family">Khalifa</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Samar</namePart>
<namePart type="family">Magdy</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zaid</namePart>
<namePart type="family">Alyafeai</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Injy</namePart>
<namePart type="family">Hamed</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Wajdi</namePart>
<namePart type="family">Zaghouani</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Imed</namePart>
<namePart type="family">Zitouni</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">979-8-89176-497-2</identifier>
</relatedItem>
<abstract>We present the L3IA Morocco team’s system for shared Task B, designed for large language models and focused on detecting harmful prompts in Arabic. In this context, we define two subtasks: B1 and B2, which refer, respectively, to the binary safe/unsafe classification and the classification of harmfulness into 7 categories. Our system combines three fine-tuned Arabic encoders: AraBERTv2, CAMeLBERTmix, and MARBERT, using soft voting that takes into account the predominant classes. It achieves a macro-F1 score of 0.946 for B1 and 0.902 for B2 during the development phase. In contrast, performance drops to 0.639 for B1 and 0.758 for B2 in the blind test. We demonstrate that this generalization divergence is due to a “unsafe” overprediction bias resulting from an unbalanced training corpus (80% unsafe) and not due to alignment of identifiers or dialects. Finally, we observed a negative result: in zeroshot and few-shot, AceGPT-7B, a generative Arabic LLM, performed poorly compared to fine-tuned encoders (0.38-0.42 on Darija).</abstract>
<identifier type="citekey">saidi-khalid-2026-l3ia</identifier>
<location>
<url>https://aclanthology.org/2026.arabicnlp-sharedtasks.42/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>338</start>
<end>346</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T L3IA Morocco at ArGuard 2026: Intent over Words - Uncovering Implicit Harm in Arabic LLM Prompts
%A Saidi, Wiam
%A Khalid, Satori
%Y Abdul-Mageed, Muhammad
%Y Elsayed, Tamer
%Y Bouamor, Houda
%Y El-Beltagy, Samhaa
%Y Al-Khalifa, Hend
%Y Abdelali, Ahmed
%Y Elmadany, AbdelRahim
%Y El Mekki, Abdellah
%Y Darwish, Kareem
%Y Alkhereyf, Sakhar
%Y Luqman, Hamzah
%Y Khalifa, Salam
%Y Magdy, Samar
%Y Alyafeai, Zaid
%Y Hamed, Injy
%Y Zaghouani, Wajdi
%Y Zitouni, Imed
%S Proceedings of The Fourth Arabic Natural Language Processing Conference: Shared Tasks
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-497-2
%F saidi-khalid-2026-l3ia
%X We present the L3IA Morocco team’s system for shared Task B, designed for large language models and focused on detecting harmful prompts in Arabic. In this context, we define two subtasks: B1 and B2, which refer, respectively, to the binary safe/unsafe classification and the classification of harmfulness into 7 categories. Our system combines three fine-tuned Arabic encoders: AraBERTv2, CAMeLBERTmix, and MARBERT, using soft voting that takes into account the predominant classes. It achieves a macro-F1 score of 0.946 for B1 and 0.902 for B2 during the development phase. In contrast, performance drops to 0.639 for B1 and 0.758 for B2 in the blind test. We demonstrate that this generalization divergence is due to a “unsafe” overprediction bias resulting from an unbalanced training corpus (80% unsafe) and not due to alignment of identifiers or dialects. Finally, we observed a negative result: in zeroshot and few-shot, AceGPT-7B, a generative Arabic LLM, performed poorly compared to fine-tuned encoders (0.38-0.42 on Darija).
%U https://aclanthology.org/2026.arabicnlp-sharedtasks.42/
%P 338-346
Markdown (Informal)
[L3IA Morocco at ArGuard 2026: Intent over Words - Uncovering Implicit Harm in Arabic LLM Prompts](https://aclanthology.org/2026.arabicnlp-sharedtasks.42/) (Saidi & Khalid, ArabicNLP 2026)
ACL