The Integration of Generative Artificial Intelligence in the Academic Writing Process A Qualitative Content Analysis of Student Reflections

Abstract

This article uses qualitative content analysis to examine the integration of generative artificial intelligence (GenAI) into the academic writing process. Based on 405 codes from 19 student reflection texts from a seminar on written language acquisition (SSE) at the University of Passau, the study identifies three characteristic development paths of GenAI literacy: exploration and experimentation, critical reflection and boundary experiences, and strategic integration and maturity. The empirical findings show a highly differentiated and strategic use of AI tools across all writing phases. In 39.3 % of the codes relating to integration, the focus is on critical-evaluative skills, conscious separation of phases between human creative work and AI-supported structuring, and systematic quality control strategies. A further 30.6 % of the codes relate to transformations in writing behaviour and 30.1 % to motivational and creative effects. The study empirically proves that, with didactic guidance, students can use GenAI not only technically but also ethically and reflectively. Critical-evaluative competence is identified as the most important emerging key competence – a direct response to the error-prone nature of AI tools. The results derive concrete didactic implications: process-oriented writing didactics, new task formats with a focus on meta-reflection and prompt documentation, and process-based assessment criteria to maintain scientific integrity.