# BanglaShield offers script‑robust safety classifier for Bengali prompts

> Researchers release BanglaShield, a fine‑tuned model that detects unsafe content in native, romanized and code‑mixed Bangla with far higher accuracy than existing LLM classifiers.

Oossa · 2026-10-20 · https://oossa.com/en/banglashield-offers-script-robust-safety-classifier-for-bengali-prompts

A team led by Md Nishat Raihan published BanglaShield on arXiv. The resource contains 10,245 Bengali prompts in three script forms, each labeled for safety. When tested on six current large language models, those models marked 47 % of unsafe native‑script prompts as safe, and 63 % when the text was romanized. BanglaShield, a compact fine‑tuned encoder, achieved 87.8 macro‑F1 on native‑script and 84.4 on romanized prompts, beating the best LLM by about 45 points while using roughly 130 × fewer parameters.

## The facts

- 10,245 Bengali prompts labeled across Safe, Offensive, Harmful, Severe categories
- BanglaShield reaches 87.8 macro‑F1 on native script, 84.4 on romanized prompts

## Why it matters

Developers of Bengali language applications can use BanglaShield to filter unsafe content more reliably than current LLM‑based classifiers.

## Sources & references

1. [BanglaShield: A Script-Robust Safety Classifier and Benchmark for Bangla Native, Romanized, and Code-Mixed Prompts](https://arxiv.org/abs/2610.24288) – arXiv, 2026-10-20

Last updated: 2026-10-20