The AraGenEval shared task on Arabic authorship style transfer and AI generated text detection

Shadi Abudalfa; Saad Ezzini; Ahmed Abdelali; Hamza Alami; Abdessamad Benlahbib; Salman Chafik; Mo El-Haj; Abdelkader El Mahdaouy; Mustafa Jarrar; Salima Lamsiyah; Hamzah Luqman

doi:10.18653/v1/2025.arabicnlp-sharedtasks.1

The AraGenEval shared task on Arabic authorship style transfer and AI generated text detection

Shadi Abudalfa
, Saad Ezzini
, Ahmed Abdelali
, Hamza Alami
, Abdessamad Benlahbib
, Salman Chafik
, Mo El-Haj
, Abdelkader El Mahdaouy
, Mustafa Jarrar
, Salima Lamsiyah
, Hamzah Luqman

Language, Culture, and Communication Department (LCCD)

King Fahd University of Petroleum and Minerals
Humain
Sidi Mohamed Ben Abdellah University
Mohammed VI Polytechnic University
VinUniversity
Lancaster University
Birzeit University
University of Luxembourg

Research output: Contribution to conference › Paper › peer-review

Abstract

We present an overview of the AraGenEval shared task, organized as part of the ArabicNLP 2025 conference. This task introduced the first benchmark suite for Arabic authorship analysis, featuring three subtasks: Authorship Style Transfer, Authorship Identification, and AI-Generated Text Detection. We curated high-quality datasets, including over 47,000 paragraphs from 21 authors and a balanced corpus of human- and AI-generated texts. The task attracted significant global participation, with 72 registered teams from 16 countries. The results highlight the effectiveness of transformer-based models, with top systems leveraging prompt engineering for style transfer, model ensembling for authorship identification, and a mix of multilingual and Arabic-specific models for AI text detection. This paper details the task design, datasets, participant systems, and key findings, establishing a foundation for future research in Arabic stylistics and trustworthy NLP.

Original language	English
Pages	1-13
Number of pages	13
DOIs	https://doi.org/10.18653/v1/2025.arabicnlp-sharedtasks.1
Publication status	Published - Nov 2025
Event	Proceedings of The Third Arabic Natural Language Processing Conference - ArabicNLP, Suzhou, China Duration: 8 Nov 2025 → 9 Nov 2025

Conference

Conference	Proceedings of The Third Arabic Natural Language Processing Conference
Country/Territory	China
City	Suzhou
Period	8/11/25 → 9/11/25

Access to Document

10.18653/v1/2025.arabicnlp-sharedtasks.1

https://scholar.google.com/citations?view_op=view_citation&hl=en&user=HiUpTkcAAAAJ&pagesize=100&citation_for_view=HiUpTkcAAAAJ:x9HjRiAMpasC

Cite this

Abudalfa, S., Ezzini, S., Abdelali, A., Alami, H., Benlahbib, A., Chafik, S., El-Haj, M., El Mahdaouy, A., Jarrar, M., Lamsiyah, S., & Luqman, H. (2025). The AraGenEval shared task on Arabic authorship style transfer and AI generated text detection. 1-13. Paper presented at Proceedings of The Third Arabic Natural Language Processing Conference, Suzhou, China. https://doi.org/10.18653/v1/2025.arabicnlp-sharedtasks.1

@conference{d36452a5bdae474689c18bfe08673a42,

title = "The AraGenEval shared task on Arabic authorship style transfer and AI generated text detection",

abstract = "We present an overview of the AraGenEval shared task, organized as part of the ArabicNLP 2025 conference. This task introduced the first benchmark suite for Arabic authorship analysis, featuring three subtasks: Authorship Style Transfer, Authorship Identification, and AI-Generated Text Detection. We curated high-quality datasets, including over 47,000 paragraphs from 21 authors and a balanced corpus of human- and AI-generated texts. The task attracted significant global participation, with 72 registered teams from 16 countries. The results highlight the effectiveness of transformer-based models, with top systems leveraging prompt engineering for style transfer, model ensembling for authorship identification, and a mix of multilingual and Arabic-specific models for AI text detection. This paper details the task design, datasets, participant systems, and key findings, establishing a foundation for future research in Arabic stylistics and trustworthy NLP.",

author = "Shadi Abudalfa and Saad Ezzini and Ahmed Abdelali and Hamza Alami and Abdessamad Benlahbib and Salman Chafik and Mo El-Haj and \{El Mahdaouy\}, Abdelkader and Mustafa Jarrar and Salima Lamsiyah and Hamzah Luqman",

note = "Cited By (since 2025): 30 M1 - Query date: 2026-08-04 22:06:44 M1 - 30 cites: https://scholar.google.com/scholar?oi=bibs\&hl=en\&cites=11811522235673185309; Proceedings of The Third Arabic Natural Language Processing Conference ; Conference date: 08-11-2025 Through 09-11-2025",

year = "2025",

month = nov,

doi = "10.18653/v1/2025.arabicnlp-sharedtasks.1",

language = "English",

pages = "1--13",

}

Abudalfa, S, Ezzini, S, Abdelali, A, Alami, H, Benlahbib, A, Chafik, S, El-Haj, M, El Mahdaouy, A, Jarrar, M, Lamsiyah, S & Luqman, H 2025, 'The AraGenEval shared task on Arabic authorship style transfer and AI generated text detection', Paper presented at Proceedings of The Third Arabic Natural Language Processing Conference, Suzhou, China, 8/11/25 - 9/11/25 pp. 1-13. https://doi.org/10.18653/v1/2025.arabicnlp-sharedtasks.1

TY - CONF

T1 - The AraGenEval shared task on Arabic authorship style transfer and AI generated text detection

AU - Abudalfa, Shadi

AU - Ezzini, Saad

AU - Abdelali, Ahmed

AU - Alami, Hamza

AU - Benlahbib, Abdessamad

AU - Chafik, Salman

AU - El-Haj, Mo

AU - El Mahdaouy, Abdelkader

AU - Jarrar, Mustafa

AU - Lamsiyah, Salima

AU - Luqman, Hamzah

N1 - Cited By (since 2025): 30 M1 - Query date: 2026-08-04 22:06:44 M1 - 30 cites: https://scholar.google.com/scholar?oi=bibs&hl=en&cites=11811522235673185309

PY - 2025/11

Y1 - 2025/11

N2 - We present an overview of the AraGenEval shared task, organized as part of the ArabicNLP 2025 conference. This task introduced the first benchmark suite for Arabic authorship analysis, featuring three subtasks: Authorship Style Transfer, Authorship Identification, and AI-Generated Text Detection. We curated high-quality datasets, including over 47,000 paragraphs from 21 authors and a balanced corpus of human- and AI-generated texts. The task attracted significant global participation, with 72 registered teams from 16 countries. The results highlight the effectiveness of transformer-based models, with top systems leveraging prompt engineering for style transfer, model ensembling for authorship identification, and a mix of multilingual and Arabic-specific models for AI text detection. This paper details the task design, datasets, participant systems, and key findings, establishing a foundation for future research in Arabic stylistics and trustworthy NLP.

AB - We present an overview of the AraGenEval shared task, organized as part of the ArabicNLP 2025 conference. This task introduced the first benchmark suite for Arabic authorship analysis, featuring three subtasks: Authorship Style Transfer, Authorship Identification, and AI-Generated Text Detection. We curated high-quality datasets, including over 47,000 paragraphs from 21 authors and a balanced corpus of human- and AI-generated texts. The task attracted significant global participation, with 72 registered teams from 16 countries. The results highlight the effectiveness of transformer-based models, with top systems leveraging prompt engineering for style transfer, model ensembling for authorship identification, and a mix of multilingual and Arabic-specific models for AI text detection. This paper details the task design, datasets, participant systems, and key findings, establishing a foundation for future research in Arabic stylistics and trustworthy NLP.

U2 - 10.18653/v1/2025.arabicnlp-sharedtasks.1

DO - 10.18653/v1/2025.arabicnlp-sharedtasks.1

M3 - Paper

SP - 1

EP - 13

T2 - Proceedings of The Third Arabic Natural Language Processing Conference

Y2 - 8 November 2025 through 9 November 2025

ER -

The AraGenEval shared task on Arabic authorship style transfer and AI generated text detection

Abstract

Conference

Access to Document

Fingerprint

Cite this