Register Shifts Break LLM Safety: A Bengali Benchmark with Culturally Grounded Harms
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 reques...
| Authors | Naymul Islam, Nusrat Jahan Lia, Shubhashis Roy Dipta et al. |
|---|---|
| Primary category | cs.CL |
| Published | 23 August 2026 |
| Last updated at source | 23 August 2026 |
Models and datasets referencing this paper
- BanglaLLM/BanglaSafeDataset
Silytics shows public metadata from arXiv and links to the original. Information may be incomplete or out of date. Last refreshed 17 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.