Paper Number

ECIS2026-2081

Paper Type

SP

Abstract

Artificial Intelligence (AI), and Large Language Models (LLM) in particular, can advance global learning by enabling cross-cultural collaboration through translation, inclusive participation, enhanced communication, and feedback. Yet benefits vary with language proficiency and cultural biases in LLM design and outputs. Non-native speakers of English often face higher cognitive load and lower output quality in low-resource languages, while AI translation tools often miss cultural nuances and reflect U.S.-centred norms. In a survey of 502 students worldwide, language proficiency and nationality predicted LLM use frequency. Our main study will test how English language proficiency and cultural background influence LLM use frequency, perceived effectiveness of LLM, and perceived learning outcomes in international contexts.

Share

COinS
 
Jun 14th, 12:00 AM

Lost In Translation? Exploring Llm Use Across Language Proficiency And Cultural Boundaries

Artificial Intelligence (AI), and Large Language Models (LLM) in particular, can advance global learning by enabling cross-cultural collaboration through translation, inclusive participation, enhanced communication, and feedback. Yet benefits vary with language proficiency and cultural biases in LLM design and outputs. Non-native speakers of English often face higher cognitive load and lower output quality in low-resource languages, while AI translation tools often miss cultural nuances and reflect U.S.-centred norms. In a survey of 502 students worldwide, language proficiency and nationality predicted LLM use frequency. Our main study will test how English language proficiency and cultural background influence LLM use frequency, perceived effectiveness of LLM, and perceived learning outcomes in international contexts.