# TukaBench: A Culturally Grounded Jailbreak Benchmark for African Languages

arXiv: https://arxiv.org/abs/2606.01322 · HTML: https://mlsanity.com/paper/2606.01322/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Placed in Jailbreaking, one of 12 directions in LLM safety and security.
- Builds on 20 of its 32 references that are in the map.
- First posted on arXiv 31 May 2026.

## Placed in

- LLM safety and security / [Jailbreaking](https://mlsanity.com/safety/jailbreaking/)

## Builds on (most cited first, 20 of 20)

- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena](https://mlsanity.com/paper/2306.05685/) (2023)
- [HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal](https://mlsanity.com/paper/2402.04249/) (2024)
- [COMET: A Neural Framework for MT Evaluation](https://mlsanity.com/paper/2009.09025/) (2020)
- [Low-Resource Languages Jailbreak GPT-4](https://mlsanity.com/paper/2310.02446/) (2023)
- [A Few Thousand Translations Go a Long Way! Leveraging Pre-trained Models for African News Translation](https://mlsanity.com/paper/2205.02022/) (2022)
- [ALERT: A Comprehensive Benchmark for Assessing Large Language Models' Safety through Red Teaming](https://mlsanity.com/paper/2404.08676/) (2024)
- [Are Large Language Model-based Evaluators the Solution to Scaling Up Multilingual Evaluation?](https://mlsanity.com/paper/2309.07462/) (2023)
- [PolyGuard: A Multilingual Safety Moderation Tool for 17 Languages](https://mlsanity.com/paper/2504.04377/) (2025)
- [Code-Switching Red-Teaming: LLM Evaluation for Safety and Multilingual Understanding](https://mlsanity.com/paper/2406.15481/) (2024)
- [Multilingual Blending: LLM Safety Alignment Evaluation with Language Mixture](https://mlsanity.com/paper/2407.07342/) (2024)
- [The Token Tax: Systematic Bias in Multilingual Tokenization](https://mlsanity.com/paper/2509.05486/) (2025)
- [AfriqueLLM: How Data Mixing and Model Architecture Impact Continued Pre-training for African Languages](https://mlsanity.com/paper/2601.06395/) (2026)
- [Boundary Point Jailbreaking of Black-Box LLMs](https://mlsanity.com/paper/2602.15001/) (2026)
- [LLMs Lost in Translation: M-ALERT uncovers Cross-Linguistic Safety Inconsistencies](https://mlsanity.com/paper/2412.15035/) (2024)
- [Do Methods to Jailbreak and Defend LLMs Generalize Across Languages?](https://mlsanity.com/paper/2511.00689/) (2025)
- [UbuntuGuard: A Culturally-Grounded Policy Benchmark for Equitable AI Safety in African Languages](https://mlsanity.com/paper/2601.12696/) (2026)
- [FinSafetyBench: Evaluating LLM Safety in Real-World Financial Scenarios](https://mlsanity.com/paper/2605.00706/) (2026)
- [Break Me If You Can: Self-Jailbreaking of Aligned LLMs via Lexical Insertion Prompting](https://mlsanity.com/paper/2601.02670/) (2026)
- [One Word at a Time: Incremental Completion Decomposition Breaks LLM Safety](https://mlsanity.com/paper/2604.25921/) (2026)

## Built on by (most cited first, 3 of 3)

- [Challenges and Recommendations for LLM-as-a-Judge in Multilingual Settings and for Low-Resource Languages](https://mlsanity.com/paper/2607.02235/) (2026)
- [SpeechJBB: Probing Safety Alignment and Comprehension in Large Audio Language Models under Code-Switched Speech](https://mlsanity.com/paper/2606.06037/) (2026)
- [Language-Specific Gaps in AI Safety Training Datasets](https://mlsanity.com/paper/2608.13695/) (2026)

## Abstract (arXiv)

Safety evaluation of Large Language Models (LLMs) remains heavily English-centric, leaving Low-Resource Languages (LRLs), particularly African ones, critically underexplored. We introduce TUKABENCH, a jailbreak benchmark for seven African languages that extends JailbreakBench (JBB) beyond direct translation through four settings: human translation of JBB prompts, English adaptation to African contexts followed by human translation, human-curated prompts validated through interactions with GPT-5.2, and code-switched prompts combining English and African languages, isolating the effect of language, cultural grounding, and prompt evasiveness on model safety. Across closed and open models, prompting in African languages reduces refusal relative to English, with culturally adapted prompts leading to least refusal. The evaluation also surfaces two structural limitations: model comprehension failures and reduced LLM-as-a-judge reliability in LRLs. To capture the first, we introduce Deflection alongside Refused and Jailbroken; to assess the second, we validate outputs with human annotations, showing that judge-human agreement drops in lower-resource languages and less commonly supported scripts.