Call for Participation: ArGuard 2026 Shared Task at ArabicNLP 2026
Firoj Alam
Dear colleagues,
We are pleased to invite researchers and practitioners to participate in ArGuard 2026: Harmful Content Detection in Arabic Memes and LLM Prompts, a shared task organized as part of ArabicNLP 2026, co-located with EMNLP 2026. ArGuard 2026 focuses on Arabic-specific challenges in harmful content detection, including dialectal variation, sarcasm, cultural references, code-switching, multimodal signals, and implicit harmful intent.
The shared task includes two tracks:
Track A: Multimodal Hateful Meme Detection
Track B: Harmful Prompt Detection for LLMs
Participants are invited to develop systems, submit predictions, and write system description papers for the ArabicNLP 2026 proceedings.
Website: https://araieval.github.io/ArGuard2026/
Data/Code: https://github.com/araieval/ArGuard-2026-tasks
Registration:
Task 1A: https://www.codabench.org/competitions/16661/
Task 1B: https://www.codabench.org/competitions/16660/
Task 2A: https://www.codabench.org/competitions/16652/
Task 2B: https://www.codabench.org/competitions/16653/
Important Dates
August 1, 2026 — Registration deadline and test set release
August 6, 2026 — Evaluation deadline
August 8, 2026 — Leaderboard release
August 18, 2026 — Shared task papers due
August 30, 2026 — Notification of acceptance
September 10, 2026 — Camera-ready papers due
October 24–29, 2026 — ArabicNLP 2026 Conference
We encourage participation from the NLP, multimodal AI, online safety, LLM safety, and Arabic language technology communities.
Please feel free to share this call with anyone who may be interested.
Best regards,
Organizers