The Impact of AI Writing Assistants on Academic Writing Performance
Yulu Cui, Weimiao He, Xingke Du, Manlin Zeng, Dongping Liu
Abstract
With the rapid rise of generative AI, writing tools are increasingly integrated into higher education, reshaping human-AI interaction and raising concerns about students' critical thinking, autonomy, and engagement. This study examined AI-assisted writing through the lens of subjectivity—students' autonomy, motivation, and self-regulation. Using a within-subjects pre-post design with 44 undergraduates, results showed that AI significantly improved writing quality, especially in linguistic precision, critical recognition, and structure and logic, while overall subjectivity remained stable. Cluster analysis revealed four usage patterns, with high-subjectivity students maintaining cognitive control and outperforming passive users. AI readiness was positively related to all subjectivity dimensions and indirectly enhanced performance via motivational and autonomy pathways. Rather than a neutral tool, AI emerges as a catalyst of educational transformation, reconfiguring student agency and compelling educators to rethink how equity and subjectivity can be preserved in AI-rich environments.