How AI Paraphrasing Tools Reshape Drafting Habits and Writing Development Among EFL University Students

The fast absorption of artificial intelligence (AI) paraphrasing tools into the academic writing institution has cast serious concerns regarding its effect on the second language academic writing. It is a qualitative phenomenological research study that examines the experiences of EFL university students in the application of AI paraphrasing tools and how the use repackages their drafting practices, writing processes, and assumes their roles in authorial agency and academic honesty. The data were gathered using the method of semi-structured and in-depth interviews with 12 undergraduate EFL students who claimed to use AI paraphrasing tools on a regular basis when composing academic texts. Using an interpretivist paradigm, the data were examined through a phenomenological thematic approach to obtain the lived experiences of the participants. This is demonstrated by the results that AI paraphrasing tools acted as both linguistic and affective scaffolds decreasing anxiety about accuracy in language and academic tone. Meanwhile, the drafting activities of students were also transformed: drafting was no longer achieved by an independent composition; instead, drafting involved rewriting, with revision now being concerned with surface rewording. The respondents have also had mixed perceptions when it comes to the development of their writing skills: some of them felt that their awareness of the language has heightened due to being exposed to AI-generated versions, but other respondents have been worried about becoming too dependent and less reliable in their ability to paraphrase without AI assistance. Also, students were actively involved in negotiating authorship and academic integrity through post-editing of AI output, but they had no certainty regarding ethical limits. The paper is an attempt to note that the role of AI paraphrasing tools in EFL academic writing is complex and ambivalent and contributes to the necessity to engage in pedagogical strategies that allow promoting reflective, self-regulated, and ethically informed AI use in higher education.