Mohamed Mahna, Rajat Pareek, Spandan Parida, Rohan Rao, Andi Hidayat Thamrin
We offer a cross-case analysis of first-person accounts by undergraduate and graduate students of their use of Large Language Models (LLMs) such as ChatGPT, that enhance their reading, writing, and learning in English-medium programs across four nations: India, Indonesia, Morocco, and the United States. The participants first wrote independent accounts of their uses of AI at school and then collectively analysed each other’s and their accounts to identify common and unique opportunities and challenges across the four contexts. Opportunities identified included assistance in reading complex English texts, conversational practice, as a study tool, and in writing as a tool for brainstorming ideas and learning to write more complex and precise sentences grammatically. In the last case, the students also noted how writing and revising prompts helped them fine-tune their writing. Two types of challenges were also noted, in which a prompt generated false citations or weak writing; and “cheating”, in which a student might risk plagiarism and detection instead of using AI as an assistive tool. In conclusion, implications for the productive, ethical use of AI programs in second-language literacy instruction are discussed. © 2025 National Association for the Teaching of English.
Faculty of Letters and Human Sciences, Mohammed V University, Rabat, Morocco; Integrated Programme in Management (IPM), Indian Institute of Management, Rohtak, India; BA (Bachelor of Arts) in Bilogical Sciences, University of Illinois at Urbana-Champaign, IL, United States; Master’s of Education (English Language), Universitas Negeri Yogyakarta, Yogyakarta, Indonesia
