Despite advances in large language model capabilities in recent years, a large gap remains in their capabilities and safety performance for many languages beyond a relatively small handful of globally dominant languages. This paper provides researchers, policymakers and governance experts with an overview of key challenges to bridging the "language gap" in AI and minimizing safety risks across languages. We provide an analysis of why the language gap in AI exists and grows, and how it creates disparities in global AI safety. We identify barriers to address these challenges, and recommend how those working in policy and governance can help address safety concerns associated with the language gap by supporting multilingual dataset creation, transparency, and research.
@article{arxiv.2505.21344,
title = {The Multilingual Divide and Its Impact on Global AI Safety},
author = {Aidan Peppin and Julia Kreutzer and Alice Schoenauer Sebag and Kelly Marchisio and Beyza Ermis and John Dang and Samuel Cahyawijaya and Shivalika Singh and Seraphina Goldfarb-Tarrant and Viraat Aryabumi and Aakanksha and Wei-Yin Ko and Ahmet Üstün and Matthias Gallé and Marzieh Fadaee and Sara Hooker},
journal= {arXiv preprint arXiv:2505.21344},
year = {2025}
}