Large Language Models (LLMs), including GPT-4, are susceptible to a cross-lingual vulnerability that compromises their safety mechanisms. This vulnerability is exploited by translating unsafe prompts from English into low-resource languages using widely accessible translation services such as Google Translate. This method effectively bypasses the LLM's safety filters, resulting in a significantly higher success rate for generating harmful responses compared to direct attacks in English. The root cause of this vulnerability is the uneven distribution of safety training data across different languages, which leads to inadequate generalization of safety protocols in low-resource languages. Mitigation steps: **For AI Developers:** * Develop and implement robust multilingual safety mechanisms that effectively generalize across various languages. * Evaluate and improve the security protocols of translation APIs when integrated with Large Language Models (LLMs). **For Model Trainers/Fine-tuners:** * Increase the diversity and quantity of safety training data to encompass a wide range of low-resource languages. * Conduct regular multilingual red-teaming exercises to identify and mitigate cross-lingual security vulnerabilities.