Exploring Students’ Use of AI Translation and Paraphrasing Tools during Academic Reading
Authors
Endang Siti Nurkholidah
Universitas Sindang Kasihh Majalengka
Shafiq ur Rehman
University of Doha For Science and Technology, Qatar
Abstract
The increasing availability of AI translation and paraphrasing tools offers new possibilities for supporting English as a Foreign Language (EFL) students’ academic reading. This issue is particularly relevant in university contexts where students are required to engage with complex English academic texts yet often encounter linguistic and conceptual difficulties. This study aims to investigate how university students use AI translation and paraphrasing tools across the before-reading, during-reading, and after-reading stages of academic reading and to examine how these stage-specific uses support students’ meaning-making processes. Using a qualitative exploratory design, data were collected from 20 university students who were familiar with AI-assisted reading practices through think-aloud protocols, reading journals, and semi-structured interviews, and were analyzed thematically. Findings reveal that students engage with AI tools in a strategic and reflective manner, employing translation for lexical support, paraphrasing for structural simplification, and both tools for reflective consolidation. Students’ use of AI was stage-specific: pre-reading translation enabled previewing and activation of prior knowledge; during reading, paraphrasing supported comprehension monitoring and verification; after reading, students consolidated understanding through summarization and clarification. The study highlights that AI tools function as both cognitive and metacognitive scaffolds, extending learners’ capacity to navigate complex academic texts while promoting learner agency and self-directed reading. These insights offer practical implications for integrating AI tools into academic literacy instruction and suggest avenues for future research on longitudinal effects, disciplinary differences, and adaptive strategies for diverse proficiency levels.