USING QUILLBOT TO ENHANCE PARAPHRASING SKILLS AMONG ALGERIAN EFL UNIVERSITY STUDENTS

Studii de gramatică contrastivă nr. 44 / 2025 144 Creative Commons Attribution-NonCommercial 4.0 International, Available at : https://studiidegramaticacontrastiva.info/home DOI: 10.5281/zenodo.18016108 USING QUILLBOT TO ENHANCE PARAPHRASING SKILLS AMONG ALGERIAN EFL UNIVERSITY STUDENTS 1 Abstract: This study investigates the effectiveness of QuillBot, an AI-powered paraphrasing tool, in enhancing Algerian EFL university students’ paraphrasing skills within linguistics coursework. Paraphrasing remains a complex academic skill for many EFL learners, especially when dealing with content-heavy, discipline-specific texts. While AI tools like QuillBot are gaining popularity, few studies have evaluated their impact through controlled interventions in subject-specific academic contexts, particularly in underrepresented regions such as North Africa. Ninety third-year English majors at Batna 2 University participated in a four-week quasi-experimental study. Students were divided into an experimental group (N = 45), which received guided instruction on using QuillBot during linguistics writing tasks, and a control group (N = 45), which followed traditional instruction. Preand post-tests measured students’ paraphrasing performance, while qualitative data were gathered through open-ended questionnaires and focus group discussions designed to explore learners’ experiences and attitudes.Quantitative findings revealed statistically significant improvements in the experimental group’s paraphrasing performance, with large effect sizes. Qualitative results showed that students valued QuillBot for expanding vocabulary, improving sentence structure, and increasing confidence in expressing complex academic content. However, concerns emerged about overreliance on the tool, limited critical reflection, and potential ethical issues when using AI-generated language. The study highlights the pedagogical potential of integrating AI tools like QuillBot into subject-specific writing instruction, particularly when combined with explicit strategies that promote critical engagement, ethical awareness, and learner autonomy. By focusing on a content-driven course in an under-researched Algerian context, this research contributes empirical evidence to the evolving discussion on AI in education and offers practical implications for instructors seeking to enhance writing instruction through responsible AI integration. Keywords: QuillBot, paraphrasing AI tools, subject-specific writing, Algerian EFL students. 1 Saida Tobbi, Université Batna 2, [email protected] Received: August 6, 2025 | Revised: October 3, 2025 | Accepted: November 19, 2025 | Published: December 22, 2025 Studii de gramatică contrastivă nr. 44 / 2025 145 Creative Commons Attribution-NonCommercial 4.0 International, Available at : https://studiidegramaticacontrastiva.info/home UTILISATION DE QUILLBOT POUR AMÉLIORER LES COMPÉTENCES EN PARAPHRASE CHEZ LES ÉTUDIANTS ALGÉRIENS EN ANGLAIS LANGUE ÉTRANGÈRE À L'UNIVERSITÉ Résumé : Cette étude examine l'efficacité de QuillBot, un outil de paraphrase alimenté par l'IA, dans l'amélioration des compétences en paraphrase des étudiants algériens en anglais langue étrangère (EFL) dans le cadre de cours de linguistique. La paraphrase reste une compétence académique complexe pour de nombreux apprenants de l'anglais langue étrangère, en particulier lorsqu'il s'agit de textes riches en contenu et spécifiques à une discipline. Alors que les outils d'IA tels que QuillBot gagnent en popularité, peu d'études ont évalué leur impact à travers des interventions contrôlées dans des contextes académiques spécifiques à une matière, en particulier dans des régions sous-représentées telles que l'Afrique du Nord. Quatre-vingt-dix étudiants de troisième année en anglais à l'université de Batna 2 ont participé à une étude quasi expérimentale de quatre semaines. Les étudiants ont été répartis en un groupe expérimental (N = 45), qui a reçu des instructions guidées sur l'utilisation de QuillBot lors de tâches de rédaction linguistique, et un groupe témoin (N = 45), qui a suivi un enseignement traditionnel. Des tests pré et post-intervention ont mesuré les performances des étudiants en matière de paraphrase, tandis que des données qualitatives ont été recueillies au moyen de questionnaires ouverts et de discussions de groupe conçues pour explorer les expériences et les attitudes des apprenants. Les résultats quantitatifs ont révélé des améliorations statistiquement significatives des performances de paraphrase du groupe expérimental, avec des effets de grande ampleur. Les résultats qualitatifs ont montré que les étudiants appréciaient QuillBot pour l'enrichissement de leur vocabulaire, l'amélioration de la structure de leurs phrases et le renforcement de leur confiance dans l'expression de contenus académiques complexes. Cependant, des inquiétudes ont été soulevées concernant la dépendance excessive à cet outil, le manque de réflexion critique et les problèmes éthiques potentiels liés à l'utilisation d'un langage généré par l'IA. L'étude met en évidence le potentiel pédagogique de l'intégration d'outils d'IA tels que QuillBot dans l'enseignement de l'écriture dans des matières spécifiques, en particulier lorsqu'ils sont associés à des stratégies explicites qui favorisent l'engagement critique, la conscience éthique et l'autonomie des apprenants. En se concentrant sur un cours axé sur le contenu dans un contexte algérien peu étudié, cette recherche apporte des preuves empiriques au débat en cours sur l'IA dans l'éducation et offre des implications pratiques pour les enseignants qui cherchent à améliorer l'enseignement de l'écriture grâce à une intégration responsable de l'IA. Mots-clés : QuillBot, outils d'IA de paraphrase, écriture disciplinaire, étudiants algériens en anglais langue étrangère. 1. Introduction Writing remains one of the most demanding skills for students learning English as a foreign language, particularly at the university level. It requires more than accuracy in grammar and vocabulary; students must learn to organize complex information, express ideas clearly, and engage critically with academic sources. These challenges intensify in disciplines like linguistics, where students are expected not only to grasp abstract theoretical concepts but also to analyze and apply them in written assignments (Rahmayani, 2018; Tran & Nguyen, 2022). In this context, paraphrasing emerges as a particularly crucial skill. It allows learners to reframe scholarly material in their own words, helping them integrate knowledge meaningfully and avoid plagiarism—yet it is often one of the most difficult skills to master. Studii de gramatică contrastivă nr. 44 / 2025 146 Creative Commons Attribution-NonCommercial 4.0 International, Available at : https://studiidegramaticacontrastiva.info/home Effective paraphrasing calls for more than linguistic substitution; it demands comprehension, reformulation, and a confident command of language. In linguistics courses, where students routinely summarize sources, rephrase complex definitions, and build arguments based on cited work, paraphrasing becomes essential. However, many EFL learners lack the vocabulary depth and syntactic control needed to do this well. As a result, they may fall into the habit of copying phrases too closely, risking unintentional plagiarism and weakening their writing quality (McInnis, 2009; Ramadhani, 2019). In Algerian universities, writing instruction is often teacher-centered, and explicit training in paraphrasing is rare—especially within subject-specific contexts like linguistics, where the intersection of language and disciplinary content adds another layer of difficulty (Badriyah et al., 2021). Recently, artificial intelligence (AI) tools have been introduced into academic writing instruction, offering new possibilities. Paraphrasing tools such as QuillBot are designed to help users rephrase sentences, vary their vocabulary, and improve grammatical accuracy while maintaining the original meaning (Kurniati & Fithriani, 2022; Nurmayanti & Suryadi, 2023). These tools may offer valuable scaffolding for students who struggle with language control, providing examples of syntactic variation and helping them avoid surfacelevel repetition. However, without proper guidance, students may misuse such tools, relying on them passively rather than using them as part of a thoughtful writing process (Rogerson & McCarthy, 2017). While the popularity of AI-based tools has grown, most existing research focuses on students’ attitudes and perceived benefits rather than their actual writing performance. For instance, Alammar and Amin (2023) found generally positive perceptions of QuillBot among EFL learners, but their study relied on self-reported data. Others, such as Truong Hong Ha (2023) and Amyatun and Kholis (2023), have investigated the tool’s role in essay writing, though these studies often lack robust designs or do not address specific disciplinary writing tasks. This study seeks to fill that gap by focusing on the use of QuillBot in the context of linguistics assignments among Algerian EFL students. Linguistics is a particularly challenging subject for learners, as it requires mastering both content-specific terminology and abstract reasoning. Writing assignments in this field often demand paraphrasing scholarly work—something students must do to show both understanding and originality. By integrating QuillBot into linguistics coursework and evaluating its effectiveness through a quasi-experimental design, this research offers both empirical insights and pedagogical reflections on the tool’s role in improving paraphrasing. The novelty of this study lies in its focus on a content-heavy subject, its application of AI tools in an underexplored North African context, and its dual emphasis on learning outcomes and learner experience. It not only asks whether QuillBot works but also explores how it can be used ethically and effectively to support academic writing in a linguistics classroom where students are often left without sufficient instructional scaffolding. The present study attempts to answer the following questions: 1. Does integrating QuillBot into linguistics coursework significantly enhance the paraphrasing performance of Algerian EFL students compared to traditional instruction? 2. How do students perceive the role and usefulness of QuillBot in supporting academic writing tasks in linguistics? Studii de gramatică contrastivă nr. 44 / 2025 147 Creative Commons Attribution-NonCommercial 4.0 International, Available at : https://studiidegramaticacontrastiva.info/home 3. What challenges and ethical concerns arise from the use of QuillBot in paraphrasing instruction among EFL learners? This study contributes to both applied linguistics and the growing field of educational AI by evaluating how a tool like QuillBot affects writing development in a subject-specific context. Rather than general essay writing, it focuses on linguistics assignments, where paraphrasing is often required but rarely taught. Using a quasi-experimental design and collecting qualitative reflections, the research bridges a methodological gap and adds localized knowledge from an underrepresented region. It also offers practical guidance for integrating AI tools into writing instruction in a way that supports learning, respects academic integrity, and meets the real needs of students navigating both linguistic and disciplinary complexity. 2. Literature Review Paraphrasing is widely acknowledged as one of the more demanding aspects of academic writing, especially for learners of English as a foreign language. It requires students not only to understand the source text but also to express the same ideas in new language while maintaining accuracy and coherence. This cognitive process places a dual burden on learners: conceptual understanding and linguistic reformulation. For many EFL learners, this process is complicated by limited vocabulary, difficulties with grammar, and a lack of experience in dealing with academic texts. Several studies have noted that poor paraphrasing skills often result in unintentional plagiarism and weaken the overall quality of student writing (Ramadhani, 2019; Rahmayani, 2018). In Algerian higher education, where explicit writing instruction remains minimal and often decontextualized, these challenges are even more acute. Writing a research paper involves not just content generation but also the ability to engage critically with existing literature. This includes summarizing and paraphrasing sources in ways that reflect understanding and avoid excessive dependence on the original language. Tran and Nguyen (2022) emphasize that paraphrasing is more than a technical skill—it reflects students' capacity for analysis and synthesis. Yet, despite its centrality in academic writing, paraphrasing is often underrepresented in formal instruction. McInnis (2009) points out that students who struggle with paraphrasing are also more likely to face difficulties in completing research assignments, particularly those that require academic integrity and original thought. These difficulties are not simply linguistic but pedagogical, stemming from a lack of sustained training in textual transformation and source integration. In recent years, the emergence of AI-powered writing tools has added a new dimension to the way students approach academic writing. Tools like QuillBot, Grammarly, and Ginger have become increasingly popular for their ability to rephrase, correct grammar, and improve fluency. QuillBot, in particular, has drawn attention because of its paraphrasing function, which offers learners a variety of sentence structures and vocabulary alternatives (Kurniati & Fithriani, 2022). This has made it appealing to students who lack confidence in their linguistic abilities or who need support in refining their writing. However, the pedagogical implications of such tools remain underexplored, especially in formal educational contexts where writing is tied to assessment and originality. Some researchers see these tools as having genuine educational value. For instance, a study by Truong Hong Ha (2023) found that Vietnamese EFL students benefited from using QuillBot, especially when preparing for IELTS writing tasks. The tool helped students to become more aware of lexical and syntactic options, and many reported that it made them Studii de gramatică contrastivă nr. 44 / 2025 148 Creative Commons Attribution-NonCommercial 4.0 International, Available at : https://studiidegramaticacontrastiva.info/home feel more confident about their writing. Similarly, Nurmayanti and Suryadi (2023) reported that Indonesian students using QuillBot demonstrated noticeable improvement in sentence complexity and organization. Yet these studies tend to focus on learner perceptions or testpreparation contexts, and they often lack empirical rigor or application to academic coursework. However, not all research presents an entirely positive view. Several scholars have raised concerns about overreliance on such tools. Alammar and Amin (2023), for example, observed that many students in their study used automated paraphrasing tools uncritically. They accepted the suggestions without reflecting on meaning or structure, which limited their learning and led to superficial improvements. There is also the risk, as Rogerson and McCarthy (2017) caution, that students may blur the line between using a tool for support and using it to complete a task on their behalf, which raises ethical issues around authorship and plagiarism. Such concerns are particularly relevant in academic environments where expectations of originality and transparency are high but not always clearly communicated to students. One recurring limitation in the current literature is the lack of robust research designs. Many existing studies rely on student self-reports or short-term observations, and few include control groups or objective performance measures. As a result, while students often say they benefit from AI tools, we know less about whether these tools lead to lasting improvements in writing ability. Moreover, most of the available research has been conducted in Asian or Gulf contexts, with little attention paid to educational settings in North Africa. This limits the generalizability of findings to Algerian universities, where students may face different linguistic, technological, and pedagogical challenges. Furthermore, there is a notable absence of studies that examine the integration of AI tools in subject-specific writing tasks, where disciplinary knowledge and specialized vocabulary introduce additional complexity. In addition to questions about effectiveness, researchers have also explored how students perceive AI tools. Studies show that learners tend to view tools like QuillBot positively, appreciating the convenience and support they offer. But perceptions can be shaped more by the ease of use than by any real impact on learning. For example, while students in the study by Truong Hong Ha (2023) liked using QuillBot, few of them reported thinking critically about the tool’s suggestions. In another study, Amyatun and Kholis (2023) found that some students relied on QuillBot not just for paraphrasing but as a way to complete tasks more quickly, without necessarily engaging in deeper revision or learning. These patterns suggest that the pedagogical potential of AI tools is contingent on how they are introduced, framed, and scaffolded by instructors. These findings point to the need for a more guided and reflective use of AI tools. If students are introduced to QuillBot within a structured instructional context—one that emphasizes critical thinking, revision, and academic ethics—then the tool can serve as a learning aid rather than a shortcut. Alammar and Amin (2023) highlight this approach in their study, where students received explicit training in the writing process and were encouraged to use paraphrasing tools only after drafting and revising manually. This kind of integrated approach appears to hold more promise than simply allowing unrestricted use of AI tools. It also shifts the focus from technological efficiency to pedagogical intentionality—recognizing that the tool’s value depends on how learners are taught to engage with it. In conclusion, although interest in AI tools for enhancing EFL writing—particularly paraphrasing—is growing, much of the current literature remains methodologically limited, Studii de gramatică contrastivă nr. 44 / 2025 149 Creative Commons Attribution-NonCommercial 4.0 International, Available at : https://studiidegramaticacontrastiva.info/home often relying on perception-based data or general writing contexts. There is a noticeable gap in research that combines rigorous experimental design with a focus on subject-specific academic writing, particularly in underrepresented educational settings such as Algeria. Moreover, few studies explore how AI tools perform when integrated into content-heavy coursework, where students must paraphrase complex and discipline-specific material. This study addresses that gap by employing a quasi-experimental design to evaluate the impact of QuillBot on EFL students’ paraphrasing performance in linguistics coursework. It also explores students’ experiences and challenges in engaging with the tool. The following section outlines the methodological framework adopted to systematically assess both the learning outcomes and the pedagogical implications of integrating QuillBot into a contentheavy academic environment. 3. Methodology 3.1. Research design and rationale This study was designed to investigate whether using QuillBot as an instructional support tool could help university EFL students improve their paraphrasing skills. To achieve this, a quasi-experimental design was adopted, involving one experimental group and one control group. This approach was chosen because random assignment of participants was not possible within the institutional context, where students are already placed into intact classes. However, by comparing two comparable groups—one receiving QuillBot-assisted instruction and the other receiving traditional instruction—it was still possible to draw meaningful conclusions about the tool’s impact. The quasi-experimental approach also allowed the researcher to maintain the authenticity of classroom instruction while observing how the integration of a digital tool like QuillBot could influence learning outcomes. Since paraphrasing is a skill that depends on practice and feedback, the design made it possible to embed the tool into actual writing tasks and monitor its effects over time. Including both quantitative and qualitative tools helped balance objective performance results with students’ personal experiences and perceptions. 3.2. Participants The study involved 90 third-year undergraduate students (aged 20–27) from the Department of English Language and Literature at Batna 2 University, Algeria. They were divided into two intact classes: an experimental group (n = 45) and a control group (n = 45). Group assignment was based on existing class divisions to maintain the regular academic schedule. At this stage in their program, students are expected to engage in research-based writing and are introduced to academic skills such as paraphrasing, summarizing, and integrating sources—making them well-suited for the study’s objectives. All participants had similar language backgrounds and writing experience and were considered upper-intermediate EFL users (approximately B2 level on the CEFR scale), based on institutional placement tests and coursework history. None had received formal instruction in paraphrasing prior to the study, though 18% of the participants reported limited informal exposure to QuillBot or similar tools. To ensure baseline equivalence, both groups completed a paraphrasing pre-test before the intervention, which revealed no statistically significant differences in initial performance. Studii de gramatică contrastivă nr. 44 / 2025 150 Creative Commons Attribution-NonCommercial 4.0 International, Available at : https://studiidegramaticacontrastiva.info/home 3.3. Instructional procedure The intervention took place over four weeks, from April 2 to April 30, 2025, during regular linguistics classes. Each session lasted about 90 minutes and was conducted by the course instructor, who also served as the researcher. To maintain the authenticity of the classroom environment and avoid disrupting the existing schedule, the intervention was embedded into the regular course activities. At the start and end of the intervention, both the experimental and control groups completed the same paraphrasing preand post-tests. The experimental group received focused instruction on how to use QuillBot to support their paraphrasing. Students were introduced to the tool in class and guided in using it specifically for rewording texts related to linguistics, including academic readings and research-based content. During class sessions, they practiced paraphrasing both with and without the tool, compared different versions of their work, and discussed the lexical and structural changes suggested by QuillBot. Weekly assignments were submitted for feedback from the instructor. The control group received conventional instruction in paraphrasing. They worked with the same materials and completed similar tasks but without the use of QuillBot or any AI-based writing tools. Their instruction relied on teacher explanation, modeling, hands-on rewriting tasks, and instructor feedback. By integrating QuillBot into actual coursework, the intervention provided the experimental group with a realistic and relevant use of the tool. This allowed the study to assess the impact of AI-assisted paraphrasing within a meaningful and discipline-specific learning context. 3.4. Data collection tools To evaluate the impact of the intervention, multiple forms of data were collected: 3.4.1. Pre-test and post-test The paraphrasing rubric used for both tests consisted of four key dimensions: (1) Accuracy of meaning transfer, (2) Lexical diversity and appropriateness, (3) Grammatical range and correctness, and (4) Degree of syntactic transformation. Each response was rated on a fivepoint scale per criterion (maximum total = 20). To illustrate, a high-quality paraphrase would rephrase a dense academic definition by changing both structure and vocabulary while preserving meaning, whereas a low-quality one would merely substitute a few words or distort the original intent. All students completed a paraphrasing test before and after the instructional period. The test required them to paraphrase academic passages, and their responses were assessed using a rubric that focused on clarity, accuracy, grammatical correctness, and lexical variation. The same criteria were used for both tests to ensure consistency. 3.4.2. Students’ questionnaire After the intervention, a questionnaire was given to the experimental group to explore their experiences with QuillBot. It included both scaled items and open-ended questions covering Studii de gramatică contrastivă nr. 44 / 2025 151 Creative Commons Attribution-NonCommercial 4.0 International, Available at : https://studiidegramaticacontrastiva.info/home ease of use, perceived benefits, challenges, and concerns about using the tool in academic writing. 3.4.3. Focus group interviews A smaller group of students from the experimental class participated in follow-up interviews. Specifically, two 45-minute focus group discussions were held with a total of 10 students. These discussions allowed the researcher to explore students’ reflections in more depth— particularly their thoughts on using QuillBot as part of their writing process. 3.5. Data analysis The test results were analyzed using both within-group and between-group comparisons. A paired-sample t-test was used to see if there was a significant improvement in each group’s performance from pre-test to post-test. An independent samples t-test compared the two groups to determine whether the gains made by the experimental group were statistically different from those of the control group. Responses from the questionnaire were summarized using descriptive statistics to highlight common patterns in student attitudes toward the tool. The interview transcripts and open-ended responses were analyzed thematically using Braun and Clarke’s (2006) sixphase framework. 3.6. Ethical considerations All participants were fully informed about the purpose and procedures of the study, and their participation was entirely voluntary. They were assured that they could withdraw at any time without consequence, and that their involvement would have no impact on their academic evaluation or course grades. To maintain confidentiality and protect student identities, no personally identifiable data were collected. All names used in the reporting of qualitative findings are pseudonyms. Both the experimental and control groups received equal instructional time, and students in the control group were later granted access to QuillBot to ensure fairness and equity in learning opportunities. 4. Results This section outlines the main findings of the study, based on both the quantitative test scores and qualitative data gathered from the experimental group. The results are presented in four parts. First, the pre-test scores are examined to ensure the two groups were comparable before the intervention. Second, changes within each group between preand post-tests are analyzed. Third, the post-test performances of the experimental and control groups are compared. Finally, students’ perceptions and reflections on their experience with QuillBot are discussed, based on questionnaire responses and focus group feedback. 4.1. Pre-Intervention group equivalence To ensure the experimental and control groups were comparable before the intervention, an independent samples t-test was conducted on their pre-test scores. The results, presented in Table 1, show no statistically significant difference between the groups, confirming a similar starting point. Studii de gramatică contrastivă nr. 44 / 2025 152 Creative Commons Attribution-NonCommercial 4.0 International, Available at : https://studiidegramaticacontrastiva.info/home Group Mean (M) SD t p Experimental 9.82 1.31 0.60 0.55 Control 9.63 1.26 Table 1. Independent Samples T-Test for Pre-Test Scores 4.2. Within-group comparisons Paired-sample t-tests were used to measure progress within each group from pre-test to posttest. Results are shown in Table 2. Group Pre-Test M (SD) Post-Test M (SD) t p Effect Size (d) Experimental 9.82 (1.31) 13.06 (1.44) 11.48 < .001 1.94 Control 9.63 (1.26) 10.71 (1.38) 4.31 < .001 0.73 Table 2. Within-Group Gains in Paraphrasing Scores The experimental group showed a larger and statistically stronger improvement than the control group. The high effect size (d = 1.94) indicates a robust impact of QuillBot-assisted instruction. 4.3. Between-group comparison of post-test scores An independent samples t-test comparing the post-test means of both groups revealed a statistically significant difference favoring the experimental group. Group Post-Test Mean (SD) t p Effect Size (d) Experimental 13.06 (1.44) 7.08 < .001 1.69 Control 10.71 (1.38) Table 3. Comparison of Posttest Scores Between Experimental and Control Groups These results confirm that students who received QuillBot-based instruction significantly outperformed their peers in the control group in terms of paraphrasing accuracy and sophistication. 4.4. Students’ perceptions of QuillBot (Experimental group only) A post-intervention questionnaire was administered to the experimental group to gather feedback on their experiences with QuillBot. Studii de gramatică contrastivă nr. 44 / 2025 159 Creative Commons Attribution-NonCommercial 4.0 International, Available at : https://studiidegramaticacontrastiva.info/home Hyland, K. (2016). Teaching and researching writing (3rd ed.). Abingdon, UK: Routledge. https://doi.org/10.4324/9781315692913 Joost, N. K. (2018). Artificial intelligence: Definitions, dimensions, and directions. In AI in Education (pp. 9–26). New York, NY: Springer. https://doi.org/10.1007/978-3-319-938462_2 Kartawijaya, S. (2018). Teaching writing descriptive text by using video to the seventh grade students of SMP Negeri 2 Gunungpelindung. Premise Journal, 7(1), 58–75. https://doi.org/10.24127/pj.v7i1.1239 Kholis, A. (2018). Writing difficulties faced by EFL students: Causes and solutions. Journal of English Education, Literature, and Linguistics, 1(2), 1–9. https://doi.org/10.31539/jeell.v1i2.246 Koltovskaia, S. (2020). Student perceptions of automated writing evaluation software. Journal of Response to Writing, 6(1), 40–66. https://journalrw.org/index.php/jrw/article/view/158 Kurniati, E. Y., & Fithriani, R. (2022). QuillBot: A digital writing assistant for EFL learners. Studies in English Language and Education, 9(2), 511–524. https://doi.org/10.24815/siele.v9i2.24961 Lin, L., & Morrison, B. (2021). The challenges of EFL academic writing in higher education: Insights from Chinese students and teachers. Journal of English for Academic Purposes, 50, 100958. https://doi.org/10.1016/j.jeap.2021.100958 Maulidina, F., & Wibowo, A. E. (2022). Students’ problems in writing: A case study in an Indonesian senior high school. JELLT (Journal of English Language and Language Teaching), 6(1), 34–43. https://doi.org/10.36597/jellt.v6i1.10465 McCarthy, J. (2004). What is artificial intelligence? AI Topics. Stanford University. http://www-formal.stanford.edu/jmc/whatisai.html McInnis, R. (2009). Avoiding plagiarism: What do students know? The International Journal for Educational Integrity, 5(2), 1–13. https://doi.org/10.21913/IJEI.v5i2.603 Mozgovoy, M. (2011). Enhancing grammar checking using machine learning techniques. International Journal of Artificial Intelligence in Education, 21(3), 325–346. https://doi.org/10.3233/JAI-2011-019 Nurmayanti, L., & Suryadi, A. (2023). QuillBot AI and its effectiveness in supporting EFL writing practices. Journal of Language and Education Innovation, 4(1), 10–24. https://doi.org/10.56003/jlei.v4i1.198 Nurfiryalianti, R., Santoso, T., & Wahyuni, S. (2014). Teaching writing by using mind mapping technique at the second grade of SMA Negeri 2 Metro. Premise Journal, 3(2), 85– 97. Popenici, S. A. D., & Kerr, S. (2017). Exploring the impact of artificial intelligence on teaching and learning in higher education. Research and Practice in Technology Enhanced Learning, 12(1), 1–13. https://doi.org/10.1186/s41039-017-0062-8 Pratama, Y. (2021). Students' perceptions on writing difficulties and their efforts to overcome them. JELLT (Journal of English Language and Language Teaching), 5(1), 65–78. https://doi.org/10.36597/jellt.v5i1.10347 Rahmayani, D. (2018). Paraphrasing in EFL academic writing: Strategies and difficulties. Langkawi: Journal of The Association for Arabic and English, 4(1), 12–20. https://doi.org/10.31332/lkw.v4i1.839 Studii de gramatică contrastivă nr. 44 / 2025 160 Creative Commons Attribution-NonCommercial 4.0 International, Available at : https://studiidegramaticacontrastiva.info/home Ramadhani, D. (2019). The role of paraphrasing in avoiding plagiarism: An overview. Journal of English Education and Teaching, 3(2), 158–165. https://doi.org/10.33369/jeet.3.2.158-165 Rogerson, A., & McCarthy, G. (2017). Using internet resources to avoid plagiarism: An insight into students’ attitudes. Active Learning in Higher Education, 18(3), 231–243. https://doi.org/10.1177/1469787417715202 Tran, T. T., & Nguyen, T. M. (2022). Paraphrasing in EFL academic writing: Difficulties and strategies. Asian EFL Journal, 24(2.1), 103–123. Truong, H. H. (2023). Using QuillBot for enhancing EFL learners’ essay writing skills. HNUE Journal of Science: Educational Sciences, 68(3), 41–50. https://doi.org/10.18173/2354-1075.2023-0060 Woodward-Kron, R. (2007). Negotiating meanings and scaffolding academic writing. Journal of English for Academic Purposes, 6(4), 234–250. https://doi.org/10.1016/j.jeap.2007.05.002 Zhang, Y. (2020). An investigation into the use of automated writing evaluation tools among Chinese college students. Computer Assisted Language Learning, 33(8), 951–972. https://doi.org/10.1080/09588221.2019.1577892 Dr. Saida TOBBI is an Associate Professor in the Department of English at Batna 2 University, Algeria. She holds a Magister in Pragmatics and Natural Language Understanding from East China Normal University (Shanghai, China) and a PhD in Linguistics from the University of Batna 2. Dr. Tobbi currently serves as the Director of the research laboratory Extended Disciplinary Competencies in English as a Foreign Language. She has published several papers in international journals and has participated in numerous national and international conferences. Her research interests include general and applied linguistics, sociolinguistics, pragmatics, discourse analysis, and cultural studies. ORCID: https://orcid.org/0009-0005-9602-3411