Artificial Intelligence (AI) is upon us as weekly news and innovations have become part of our routine. Many have argued that there is no escaping, as AI is here to stay. Therefore, the future is for those who know how to work with AI. In the field of education, and particularly educational technology, researchers have explored the potential of AI models such as ChatGPT -Conditional Generative Pre-trained Transformer or ChatGPT (OpenAI, 2023) for learning and teaching (Adeshola & Adepoju, 2023). For instance, Memarian and Doleck (2023) indicate that ChatGPT can facilitate the development of personalized and complex learning, specific teaching and learning activities, assessments, asynchronous communication, and feedback, accuracy in research, personas, task delegation, and cognitive offload. Kasneci et al. (2023) add that ChatGPT can also be used to semi-automate the grading process for students’ work by identifying both the strengths and weaknesses within a given activity in a broad range of assignments, including research articles, academic essays, and other forms of written coursework. Research further shows that ChatGPT can be used for English language learning. Phuong (2024) finds that students in Korea extensively used ChatGPT as a learning assistance tool to solve English tasks and assignments. In fact, students predominantly used ChatGPT to find instant solutions to English learning difficulties. For learners of English as a foreign language (EFL), ChatGPT can improve learners’ language proficiency and basic language skills of speaking, listening, reading, grammar, vocabulary, and particularly writing skills (Balcı, 2024). Given the educational potential of ChatGPT, Grassini (2023) argues that university education should prepare students with the necessary skills to thrive in an AI-dominated future. Therefore, the integration of AI-powered technologies such as ChatGPT in educational settings is a necessary step. By providing hands-on experience with these tools, institutions can enhance students’ understanding and application, while sharing their limitations and keeping pace with technological advances. In this paper, we were interested in understanding how international students are leveraging AI to improve their English.
This qualitative study adopted a text-based vignette approach to explore how international graduate students used AI tools such as ChatGPT to learn and improve their English. Vignettes, in research contexts, are written, visual, or oral stimuli, based on realistic and identifiable situations and settings that resonate with participants to invite responses not limited to beliefs, emotions, reflections, practices, and experiences (Skilling &Stylianides, 2020). As a data collection technique, vignettes lend themselves well to this study because they allow us to capture participants’ practices and experiences with ChatGPT. Given that we were interested in participants' subjective experience, with a specific focus, without including other aspects of participants’ history (Flick, 2023), vignettes were the appropriate tool to lessen some of the drawbacks of semi-structured interviews (Bain, 2023). We employed a written vignette that involved a real-life story of international students who are in their graduate studies in the US, having English as a second or third language from their various countries. While some faculty allowed students to use AI, others forbade its use, and some international students, due to where they come from, are either faced or forced to use it secretly for several reasons. This vignette was intended to elicit participants’ responses within a scenario controlled by the researchers. Participants’ responses were recorded by the researchers, then transcribed. Braun and Clarke's (2006) thematic analysis approach to data analysis was adopted to make sense of the data. Participants were recruited using purposive sampling.
Three major themes emerged from our analysis of the data that demonstrate how participants were using ChatGPT to learn and improve their English.
As a graduate student, one participant explained that they used ChatGPT to restructure their thoughts and support their learning. They considered the tool an assistive companion for learning and reported using it frequently, sometimes daily. After drafting essays, they often uploaded their writing into ChatGPT to help reorganize ideas. When unfamiliar with a topic, they used the tool to generate preliminary ideas that guided their learning path.
Although ChatGPT provided immediate feedback, the panelist noted that they did not rely on it entirely because it occasionally produced inaccurate responses. Still, the immediacy of feedback encouraged regular use. Even when the responses did not fully align with their thinking, they offered directions on what to explore next. For this panelist, ChatGPT functioned as a learning and research assistant that, when personalized, helped shape and guide the learning process.
Another participant addressed what it meant for international students to use ChatGPT as part of their learning process. Graduate students were often reminded that AI tools would not replace them anytime soon, as learning and instruction remain highly personalized and human-centered. However, as these students entered their careers, they would need to understand how to appropriately identify, select, and utilize AI tools. AI literacy has become essential rather than optional. Examples shared included using ChatGPT to generate ideas, create visuals, refine explanations, and address research questions. Drawing on personal experience, the panelists discussed how ChatGPT supported academic writing for graduate students for whom English was a second language, helping them check grammar, improve clarity, and express ideas more effectively.
The final theme focused on how generative AI influenced the development and communication of ideas. ChatGPT assisted with initial concept development, proposing syntheses of ideas, drafting outlines, and generating visuals. Participants acknowledged the demanding nature of graduate school and emphasized that international students often faced additional challenges, such as understanding academic articles, framing ideas, and expressing thoughts in a way that instructors and native English-speaking peers could clearly understand. They shared various ways they had used ChatGPT to develop their ideas, summarize key arguments from readings, and rewrite content for class projects and discussions. Through these examples, the panelists demonstrated how generative AI could support intellectual engagement, idea formation, and communication within graduate education.
This study demonstrated that ChatGPT and related AI tools have already begun to reshape educational practices, particularly for international graduate students navigating demanding academic environments. AI’s transformative potential, ranging from personalized instruction to enhanced engagement and improved academic efficiency, signals that these technologies are becoming unavoidable components of modern learning. As generative AI and large language models continue to evolve, their implications for teaching, learning, and assessment become increasingly significant.
Drawing on our collective experiences, we highlighted the various ways ChatGPT supported students’ self-regulated learning, strengthened and sustained learning habits, and assisted in the organization and expression of complex ideas. We also shared teachers’ perspectives on how ChatGPT can complement instructional practices, streamline assessment processes, and offer timely feedback that advances student understanding.
Ultimately, this panel illustrated that ChatGPT can address a broad spectrum of student needs—from grammar refinement and idea synthesis to conceptual clarification and research support. While challenges and limitations remain, the experiences presented underscore that AI, when used thoughtfully and ethically, can enrich academic life and expand learning opportunities for diverse student populations.
EdTech Archives