HARNESSING AI FOR IDEA GENERATION: PRE-SERVICE TEACHERS’ EXPERIENCES USING CHATGPT IN ACADEMIC WRITING

Abstract

The integration of artificial intelligence (AI) in education is rapidly changing how students generate, evaluate, and develop ideas in academic writing. This qualitative study examines the cognitive processes of 10 foundation-level TESL pre-service teachers’ use of ChatGPT for idea generation with a specific focus on its influence on critical thinking. Grounded in Cognitive Load Theory, the study conceptualises ChatGPT as a cognitive offloading tool that may both support and constrain higher-order thinking during the early stages of writing. Participants used a standardised prompt to generate ideas for self-selected essay topics, followed by a critical evaluation of AI-generated content before composing their essays using the TEEL structure. Data were collected through interviews, student essays, and written reflections and analysed thematically. Findings indicate that while ChatGPT reduced cognitive load associated with idea generation and facilitated initial engagement with writing tasks, the use of ChatGPT also introduced risks of limited depth of cognitive processing, over-reliance, and uncritical acceptance of generated content. Significantly, variation was observed in how participants engaged in evaluating, selecting and refining AI-generated ideas, highlighting differences in critical thinking practices. The study contributes new insight into how AI-mediated cognitive offloading shapes idea development in academic writing, demonstrating that the effectiveness of ChatGPT depends not merely on access to the tool but on students’ ability to critically examine and transform its outputs. These findings emphasize the need to position critical thinking as a central component in AI-supported academic writing contexts.