For years, writing has been one of the biggest struggles for students learning English as a foreign language. It’s not just about grammar mistakes or awkward vocabulary choices. Writing requires students to organize ideas, connect arguments, and express thoughts clearly in another language. For many students, especially those outside language-related majors, this feels exhausting. Civil engineering students, for example, often see writing as something distant from their technical world. But in reality, academic and professional success today depends heavily on communication skills.
One of the hardest parts of academic writing is paraphrasing. Sounds simple in theory, but in practice, many students struggle to rewrite ideas without copying directly from the source. This usually happens because they lack vocabulary, confidence, or understanding of sentence structure. As a result, students often end up unintentionally plagiarizing or producing weak academic writing. Studies have shown that paraphrasing is closely tied to comprehension and critical thinking skills (Ayton et al., 2022). So when students fail at paraphrasing, the problem is often deeper than just language ability.
The challenge becomes even bigger in English for Specific Purposes (ESP) classes. Civil engineering students are not just writing random essays about hobbies or travel. They are expected to produce technical reports, explain project designs, and discuss engineering concepts professionally. That means their writing needs to be accurate, formal, and discipline-specific. General English classes rarely prepare students for this kind of communication. And honestly, many students feel lost trying to bridge the gap between technical knowledge and academic English.
This is why AI tools have suddenly become part of the conversation in education. Over the last few years, applications powered by artificial intelligence have started changing the way students approach writing. Instead of waiting for teacher feedback days later, students can now receive instant suggestions in seconds. AI tools can recommend better vocabulary, fix grammar mistakes, and even restructure sentences automatically. Naturally, this is not only creates excitement but also controversy. Some people think AI will make students lazy, while others believe it could completely transform learning.
One tool that has gained huge popularity among students is QuillBot. If you’ve spent any time around university students recently, chances are you’ve heard someone mention it. QuillBot is basically an AI-powered paraphrasing tool that rewrites sentences in different styles and structures. Students can paste a sentence, click a button, and instantly receive alternative versions. It sounds almost too easy, which is exactly why educators have mixed feelings about it. But the bigger question is not whether students use AI - it’s how they use it.
The article you wrote explored exactly this issue by focusing on civil engineering students in an ESP context . The study used a quasi-experimental design involving two groups of students. One group learned writing with the help of QuillBot, while the other relied on traditional instruction without AI assistance. Both groups completed pre-tests and post-tests measuring paraphrasing, grammar, vocabulary, and coherence. The results turned out to be surprisingly clear. Students who used QuillBot improved far more significantly than those who didn’t.
The numbers alone tell an interesting story. The experimental group’s average score jumped from 63.69 to 79.68 after using QuillBot-assisted learning. Meanwhile, the control group only improved slightly, from 62.98 to 67.69. That difference is difficult to ignore. In educational research, improvements that large usually signal something meaningful happening in the learning process. It suggests that AI tools can actually support learning instead of simply doing the work for students.
What’s especially interesting is where the improvement happened. Vocabulary development showed one of the strongest gains. Students became more comfortable using varied expressions and alternative wording in their writing. This makes sense because QuillBot constantly exposes users to multiple sentence possibilities. According to Amara et al. (2025), this kind of interaction can help learners expand their lexical repertoire naturally. In other words, students aren’t just copying. They’re being exposed to new language patterns repeatedly.
Grammar also improved noticeably. Many students struggle because they know what they want to say, but they don’t know how to structure it correctly in English. QuillBot acts almost like a silent writing tutor by showing grammatically improved alternatives. Over time, students begin recognizing correct patterns and internalizing them. This reflects what second-language researchers often call the “noticing effect,” where learners improve by repeatedly seeing accurate language input (Dai & Liu, 2024). Sometimes seeing the correction instantly is more powerful than reading long explanations in a textbook.
Another major improvement appeared in coherence and organization. Before the treatment, many students wrote paragraphs that felt disconnected or difficult to follow. After using QuillBot, their writing became more structured and easier to understand. The tool indirectly helped students organize ideas more logically. Instead of jumping randomly between thoughts, students learned how smoother sentence flow works. That’s incredibly important in technical writing, where clarity matters as much as content itself.
What makes this even more relevant is the context of civil engineering education. Engineering students are often overloaded with technical courses, calculations, and project work. Writing classes can easily feel secondary or unimportant. But modern engineering careers require communication constantly—reports, proposals, emails, presentations, and international collaboration. Hyland (2019) argues that ESP instruction should focus on real communication demands in professional contexts. So improving writing skills isn’t just an academic exercise anymore; it’s career preparation.
Still, we can’t pretend AI tools are perfect. One of the biggest concerns is dependency. If students rely too much on QuillBot, they may stop thinking critically about language choices. Instead of learning how to paraphrase independently, they might simply press a button and accept whatever appears on the screen. That creates a dangerous habit. Technology should support thinking, not replace it entirely.
There’s also the issue of academic integrity. Critics argue that AI paraphrasing tools blur the line between assistance and dishonesty. If students rewrite texts using AI without truly understanding the content, are they really learning? This concern is valid. Writing is supposed to reflect comprehension, not just linguistic manipulation. Educators therefore need to guide students carefully so that AI becomes a learning companion rather than a shortcut around learning.
At the same time, blaming AI alone feels unfair. Students have always searched for easier ways to complete assignments. The real responsibility lies in how institutions and teachers integrate technology into education. Instead of banning tools like QuillBot outright, educators could focus on teaching students ethical and meaningful ways to use them. AI literacy might actually become one of the most important academic skills in the future. Knowing how to collaborate responsibly with technology could matter more than pretending technology doesn’t exist.
Interestingly, the findings from your article also support earlier research about AI-assisted writing. Mohammad et al. (2023) found that students became more enthusiastic about writing when using QuillBot. Syahnaz and Fithriani (2023) similarly reported positive student perceptions regarding AI-assisted paraphrasing tools. Students often feel less anxious when they have immediate support during the writing process. Confidence, after all, plays a huge role in language learning success.
But what makes your study stand out is that it goes beyond perception. Many studies only ask students whether they “like” using AI tools. Your research actually measured improvement through tests and statistical analysis . That makes the findings stronger and more practical. It shows measurable progress rather than just opinions or feelings. And honestly, that kind of evidence matters when discussing technology in education.
The bigger takeaway here is that AI probably isn’t the enemy of education people fear it is. Tools like QuillBot don’t automatically destroy creativity or critical thinking. In many cases, they can reduce frustration and make learning more accessible. Students who previously struggled to express ideas may finally feel capable of writing confidently. Of course, guidance and balance remain essential. But completely rejecting AI could mean ignoring tools that genuinely help students grow.
At the end of the day, maybe the goal of education should not be protecting old methods at all costs. Maybe the goal should be helping students communicate better, think more clearly, and learn more effectively. If AI tools can support that process responsibly, then they deserve a place in modern classrooms. The challenge now is not whether AI belongs in education. The real challenge is figuring out how to use it wisely without losing the human side of learning.
References:
Amara, R., Rahmah, L. S., & Ruswandi, R. (2025). The University EFL Students’ Perceptions of QuillBot as an AI-Based Tool for Enhancing Paraphrasing Skills and Academic Writing in English. Journal of Educational Sciences, 9(5), 4298–4307. https://doi.org/10.31258/jes.9.5.p.4298-4307
Ayton, D., Hillman, C., Hatzikiriakidis, K., Tsindos, T., Sadasivan, S., Maloney, S., & Illic, D. (2022). Why do students plagiarise? Informing higher education teaching and learning policy and practice. Studies in Higher Education, 47(9), 1921–1934. https://doi.org/10.1080/03075079.2021.1985103
Dai, K., & Liu, Q. (2024). Leveraging artificial intelligence (AI) in English as a foreign language (EFL) classes: Challenges and opportunities in the spotlight. Computers in Human Behavior, 159, 108354. https://doi.org/10.1016/j.chb.2024.108354
Hyland, K. (2019). English for specific purposes: Some influences and impacts. In Second handbook of English language teaching (pp. 337–353). Springer. https://doi.org/10.1007/978-3-030-02899-2_19
Mohammad, T., Alzubi, A. A., Nazim, M., & Khan, S. I. (2023). EFL Paraphrasing Skills with QuillBot: Unveiling Students' Enthusiasm and Insights. Journal of Pedagogical Research, 7(5), 359–373. https://doi.org/10.33902/JPR.202324645
Syahnaz, M., & Fithriani, R. (2023). Utilizing artificial intelligence-based paraphrasing tool in EFL writing class. Scope: Journal of English Language Teaching, 7(2), 210–218. http://dx.doi.org/10.30998/scope.v7i2.14882
