Imagine asking a conversational bot like Claude or ChatGPT a legal question in Greek about local traffic regulations. Within seconds, it replies in fluent Greek with an answer based on UK law. The model understood the language, but not the jurisdiction. This kind of failure illustrates the inability of large language models (LLMs) to understand regional, cultural and, in this case, legal knowledge, while at the same time being proficient in many of the world’s languages.
Beyond translation: Multilingual benchmark makes AI multicultural
