Exploring multilingual international doctoral students' experience of using generative artificial intelligence for academic writing

Universities in the United States have been a popular choice for many international doctoral students worldwide. They make up a significant share of graduate program enrollment in U.S. higher education. Many international doctoral students in the United States are multilingual and use English as a second or additional language for academic writing, which can make it difficult for them to keep up with their native English-speaking peers and thrive in doctoral programs. Meanwhile, generative AI (GenAI) technology has been attracting attention for its intellectual capabilities and is being applied by people across professional fields. Researchers have recognized its potential to support students in higher education. GenAI, powered by large language models, is designed to generate human-like text in response to user goals. Given the linguistic challenges multilingual international doctoral students face, incorporating GenAI into the writing process represents a practical and promising way to support their academic writing and enhance success in doctoral programs. This qualitative multiple case study explores how multilingual international doctoral students (MIDS) use GenAI to support academic writing in a U.S. doctoral context. Guided by writing process theory, it draws on the four participants’ experiences to examine how GenAI interacts with multilingual writing practices to assist academic writing, as well as the ethical concerns that accompany GenAI's involvement in academic writing. The study offers insights into how MIDS integrate GenAI into their writing processes in academia and how such practices inform their ethical and policy considerations.