AI as a thinking partner? Exploring international students’ use of GenAI to support their academic English needs

Despite meeting university English proficiency requirements, international students who speak English as a second language often encounter difficulties with academic writing, reading, presentations, class participation and oral comprehension (e.g., Wang et al., 2023). Generative Artificial Intelligence (GenAI) tools such as ChatGPT can be particularly useful in addressing the academic language challenges of this student cohort (e.g., Han, 2024), serving as a ‘thinking partner’ and helping them fulfil the expectations to communicate proficiently in academic English (Tseng & Warschauer, 2023). However, despite GenAI’s potential and growing research in education, little is known about how international students have used this technology for academic reading, writing, listening and speaking. In addition, while self-regulated learning (SRL) is pivotal for academic success (Zimmerman, 2000), no research has examined this cohort’s use of GenAI through this lens - an emerging theoretical perspective within SLA and L2 studies (Teng & Zhang, 2022). This mixed-methods study aims to explore how international undergraduate students with varying English proficiency levels at an Australian university utilise GenAI for academic language support. Results from questionnaires and semi-structured interviews show that the participants frequently use GenAI to support their academic reading and writing, primarily due to their low confidence in their English skills and drive to meet academic expectations. Less frequent uses include preparing presentations and summarising audio content. While participants employ GenAI to monitor and complete academic tasks (SRL monitoring and control), they use it less for planning and reflection (SRL forethought and reaction/reflection). These findings suggest that international students need more scaffolded support and clearer policies on GenAI use to better self-regulate their academic language needs and maximise their capabilities and academic success. This study contributes to discussions on inclusive CALL, highlighting both opportunities and challenges in integrating GenAI into higher education.