Bengali is the seventh-most-spoken language globally, yet LLM safety evaluation remains overwhelmingly English-centric. We introduce BanglaSafe, a benchmark of 879 Bengali prompts combining 309 natively authored prompts with 570 expert-reviewed prompts, spanning 17 culturally grounded harm categories and five prompting conditions that vary language, writing style, and authority framing. Evaluating 18 frontier LLMs, we find that over half of all responses are unsafe or partially unsafe (53.6%) while 14.7% contains strictly harmful content, and that the strongest observed effect is not the switch from English to Bengali but the choice of writing style within Bengali: the same harmful request phrased as a formal newspaper investigation succeeds 17 percentage points more often than the same request phrased as a casual message, with no adversarial engineering involved. We further show that existing safety classifiers struggle to reliably evaluate Bengali content, with even frontier models failing on nearly half of all cases.
IndicSafeEval, a persuasion-based jailbreak evaluation framework for Indian languages, is introduced and it is observed that different risk categories exhibit different levels of vulnerability, with some types of harmful content being significantly more susceptible to persuasion-based jailbreaks than others.
Saikat Mondal, Mamta Mamta, Deeksha Varshney et al.· 0 citations
The Arabic Safety Index (ASAS) is introduced, the first fully human-curated Arabic benchmark for redteaming LLMs and provides a culturally grounded benchmark and redteaming protocol to drive progress in Arabic LLM safety.
F. Abed, Haidar Khan, M Saiful Bari et al.· 0 citations
SurakshaEval is introduced, a novel safety benchmark composed of human-written prompts spanning real-world scenarios, explicitly designed for ten major Indian languages - Assamese, Bengali, Gujarati, Hindi, Kannada, Malayalam, Marathi, Punjabi, Tamil, and Telugu - along with English.
Debopriyo Banerjee, K. R. Kavitha, Angana Borah et al.· 0 citations
Large language models (LLMs) are increasingly used in cross-cultural communication, content generation, intelligent question answering, and social collaboration. However, existing safety evaluations mainly focus on explicit risks such as illegality, violence, privacy leakage, hate speech, and malicious code, while payi...
Urdu, the world's tenth most spoken language with 246 million speakers, remains almost entirely absent from mainstream LLM safety evaluation and nine years of WOAH proceedings. To investigate whether this absence has measurable consequences for content moderation reliability, five large language models, GPT-4o, Claude...
F. Kara-Isitt, Sonal Khosla, S. Swift· 0 citations
CC-MMD 2026 Task B evaluates cross-cultural misogynistic meme detection. We use a fixed zero-shot VLM, Qwen2.5-VL-7B-Instruct, and vary the cultural framing of the prompt. On the English test split, positive prediction rates shift from 60.2% under a generic prompt to 61.9% under an Indian prompt, 55.0% under a Chinese...
Yu-Meng Chen, A. Zubiaga· Proceedings of the 28th Inte...· 1 citation
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
