Algorithmic Mediation and Student Creativity: Large Language Models, Academic Writing, and the Convergence of Ideas in Irish Higher Education

Abstract

The growing use of large language models in higher education has intensified debate about their influence on student writing and creative development. Existing research has often focused on the practical value of these systems for brainstorming, drafting, and improving efficiency in academic tasks. Less attention has been given to whether regular reliance on algorithmic support affects students’ confidence in generating ideas independently or contributes to greater similarity in written work. This study examines these questions through a quantitative survey of 150 undergraduate and postgraduate students from three Irish institutions: University College Cork, the University of Galway, and Maynooth University. Participants came from a range of disciplinary backgrounds, and the sample was broadly balanced by gender, comprising 78 female and 72 male students. The survey explored patterns of generative AI use, perceptions of creative self-efficacy, and student views on similarity in academic ideas and written work. The findings indicate a mixed pattern. Many respondents associated ChatGPT with faster idea generation, improved productivity, and useful support during drafting. At the same time, a notable proportion reported greater difficulty developing ideas without assistance and expressed concern that their work was becoming more similar to that of other students using the same tools. Overall, 68.7 per cent believed that their ideas had become more similar to those of other ChatGPT users, while 72.0 per cent reported that ideas they had initially regarded as original later appeared to reflect common AI-generated suggestions. These findings suggest that the educational benefits of AI-assisted writing may coexist with concerns about originality, cognitive dependence, and reduced diversity in academic expression. The study contributes to current debates on cognitive offloading, standardisation in academic writing, and creative development in AI-mediated learning environments, and points to the need for pedagogical approaches that preserve independent thinking and intellectual diversity.