Generative AI and Language Diversity: Implications for Teachers and Learners

Abstract

In this position paper, we explore the intricate relationship between English language ideologies and artificial intelligence (AI), particularly focusing on large language models (LLMs) and their implications for education. The primary objective is to examine how AI perpetuates linguistic hierarchies and reinforces imperialist and capitalist values through its reliance on standard English datasets. First, we critically outlined the impact of these technologies on linguistic diversity, academic integrity, and cultural representation in language learning. Then, we analyze the existing literature to uncover the ways in which LLMs and AI-based educational tools privilege certain varieties of English while marginalizing others. Finally, we discuss how these technologies amplify linguistic imperialism by devaluing non-standard English varieties and other languages, thereby perpetuating inequities in education. Recommendations include promoting transparency in AI training datasets, incorporating linguistic diversity in AI development, and equipping educators and learners with the tools to engage with AI ethically. This work underscores the urgent need for inclusive and equitable approaches to language education in the age of AI.

Keywords: Artificial intelligence, artificial intelligence ethics, diversity, English language ideology, large language models