By Josianne Block
When I started teaching academic writing last year, I expected Generative AI (GenAI) to be a small part of the classroom. However, I quickly realised GenAI wasn’t just a side tool; it was central to how students approached brainstorming, researching, and writing.
This shift raised several questions about my profession, with the most fundamental being: Should we still be teaching academic writing when AI can generate whole texts in seconds?
These are five things I learned in the process of attempting to answer that.
- AI can transform good writers into better ones
I use GenAI when I’m writing and in several stages of research — it saves me a lot of time. Before the rise of GenAI, I would reflect on an idea over weeks and discuss it with others, if they had the time. Now, I have that same conversation with a Large Language Model (LLM) and I improve on my idea based on its responses. It’s a back-and-forth process that is faster and feels more comfortable because GenAI does not pass judgment on me or what I type. It also saves me time during the actual writing process when it comes to modifying language inaccuracies.
However, before becoming AI-literate, I was already a good writer. Knowing what good writing involves means I can now use AI without losing my voice as a writer. I am aware of what makes writing effective, which rhetorical strategies to use, how to use GenAI’s suggestions, and when to discard them. As Jin et al. (2025) report, GenAI use which is guided by strong self-regulation strategies can improve writing quality, critical thinking, and motivation. In other words, skilled writers actually benefit the most.
However, writing is not just about generating words: it’s about making sense of our thoughts and ideas. So, while I don’t discourage the use of GenAI with my students, I do discourage over-reliance. Opting for a few ‘traditional’ lectures during which students cannot use any digital tools might help them develop their skills as writers first. Becoming a good writer means that GenAI can transform you into an even better one — but you have to be good before you can be better.
- Writing means more than using correct language
A GenAI-generated piece of academic writing may be even better, in terms of grammar and register, than what most first-year university students can independently produce. If we assume that our students are submitting work that has had the benefit of GenAI’s attention, assessing writing purely on accuracy no longer makes sense. We need to focus more on other aspects of this skill. Writing is ultimately about making choices: whether to simplify, to use complex vocabulary, or to determine the direction of a text. This is not something new, and in fact, Flower and Hayes (1981) defined writing as a series of nested decisions, with each choice influencing subsequent ones and shaping meaning.
Yet, I’ve found that most students are unaware of the decisions they make during the writing process. Educators need to make these choices more explicit by constantly asking questions like any good editor might: What’s your main argument? What is the point of this paragraph? How are you making your presence evident as an author, even if implicitly? Are you repeating yourself? Are you concluding this thought before moving on to the next one? This helps students reflect on their role as writers and the dialogic nature of writing, considering the reader’s perspective (Hyland, 2001). In doing so, we focus less on language and more on a more complex and holistic writing process. GenAI can help us with the language accuracy, and it can check for coherence, but the writer needs to do as much of the heavy lifting as possible.
- Critical reading fuels better writing
University students are constantly exposed to academic texts. These have passed through the editorial gates — in most cases, they will have been peer reviewed and will have passed through several rounds of editing. However, exposure to AI-generated content has also increased. We need to help students develop skills that critically challenge what is written. When working with academic texts, students should learn to unpack what they read by asking questions. Reading is never neutral: two people may read the same content but interpret it differently, influenced by prior knowledge, experiences, and context.
When it comes to AI-generated content, students should be encouraged to put their powers of analysis to work, too. This includes identifying the flaws, biases, and areas for improvement in that AI-generated text. On the one hand, the reader needs to ask whether what they are reading is true, or merely the product of an AI hallucination. On the other, the reader can ask themselves what they would have done differently if they had written the text themselves. This not only means developing critical thinking skills but adopting a human-centred mindset for effective and judicious use of GenAI (Xerri, 2025). Given that editing skills have become just as important as writing skills, being a good writer also means shaping and refining thoughts in a critical manner — and acting like the editor of an AI-generated text is a great way to develop the necessary skills.
- GenAI is to be used ethically and transparently
With GenAI being such a powerful tool, ethical considerations matter more than ever.
A study conducted by Alkamel and Alwagieh (2024) showed that students felt more fluent and accurate when using AI assistance, but ethical concerns remain central, as there is a major difference between outright plagiarism and using GenAI as a learning tool. There is a danger that students regard plagiarism as being equivalent to copying the work of another person, and that the use of AI therefore is some kind of clever loophole waiting to be taken advantage of.
Transparency is key and educators need to normalise the idea that using GenAI is acceptable — if done ethically, and only when that use is permitted. One way to address this is by having students share how they used the tool and what they did with its output. They could do so by sharing a link to their conversation thread with the LLM. This also gives insight into the prompts they are using and possibly how we could help them improve these. Adopting a holistic view of education means acknowledging that how students use GenAI will impact their academic performance. How we, as educators, approach GenAI in the classroom will have a direct effect on students’ behaviour when using such tools.
- The goal is building students’ confidence in navigating the 21st century
A student told me, “You’re here to teach writing. But if AI can do it better than I ever will, why should I even bother to write?”
The question stayed with me. Because GenAI can write fluently and accurately, and if you feed it the right prompts, it’ll give you a strong academic text. Hence, our role now goes beyond boosting students’ writing confidence, because pretending humans always write better than AI would be disingenuous.
Instead, we must help students become better learners — more effective in tackling assignments and more prepared for life beyond university. This starts with accepting that things have changed. After all, even the model texts we use in class might now be GenAI-assisted. One activity is comparing academic papers written pre-GenAI to those post-GenAI. This way, students could see what works, what doesn’t, and what approach they might want to adopt for their own writing. This can boost their confidence by offering them concrete examples and shifting the focus from writing to making writing choices.
Conclusion
Integrating GenAI tools into writing workflows enhances productivity and confidence, especially in creative and academic writing contexts (Li et al., 2024). This, however, requires a shift in mindset. Many university students see academic writing as something done only for assignments. We need to embed writing more holistically and connect it to other skills. One way we can do so is by incorporating multimodal resources such as videos and presentations not only during lectures but even as means of assessment (Keran, n.d.). Like writing, multimodal composition also involves rhetorical choices, and this way, we can help students become confident individuals capable of engaging with both academic and real-world challenges.
Chiang (2024) says, “It’s by living our lives in interaction with others that we bring meaning into the world. That is something that an auto-complete algorithm can never do.” Writing is one form of interaction during which conscious and unconscious choices are made. So, to my own question, ‘Should we still be teaching academic writing in the GenAI era?’ Yes, absolutely. Because teaching writing is not about resisting technology, but shaping knowledge and who we become.
References
Alkamel, M. A., & Alwagieh, N. A. (2024). Utilizing an adaptable artificial intelligence writing tool (ChatGPT) to enhance academic writing skills among Yemeni university EFL students. Social Sciences & Humanities, 10, 101095. https://doi.org/10.1016/j.ssaho.2024.101095
Chiang, T. (2024, August 31). Why A.I. isn’t going to make art. The New Yorker. https://www.newyorker.com/culture/the-weekend-essay/why-ai-isnt-going-to-make-art
Flower, L., & Hayes, J. R. (1981). A cognitive process theory of writing. College Composition and Communication, 32(4), 365–387.
Hyland, K. (2001). Bringing in the Reader: Addressee Features in Academic Articles. Written Communication, 18(4), 549–574. https://doi.org/10.1177/0741088301018004005
Jin, F., Lin, C., & Lai, C. (2025). Modeling AI-assisted writing: How self-regulated learning influences writing outcomes. Computers in Human Behavior, 165, 108538. https://doi.org/10.1016/j.chb.2024.108538
Keran, M. K. (n.d.). Designing multimodal writing assignments. In E. Hall & A. Letak (Eds.), Communicating across the curriculum: Resources for high-impact teaching with writing at UW-Madison. https://wisc.pb.unizin.org/wacsourcebook/
Li, Z., Liang, C., Peng, J., & Yin, M. (2024). The value, benefits, and concerns of generative AI-powered assistance in writing. Human-Computer Interaction, 1–25. https://doi.org/10.1145/3613904.3642625
Xerri, D. (2025, January). Critical thinking as a cornerstone of a human-centred mindset towards AI use. Learning Technologies SIG Newsletter, 9–12.
Author Biography
Josianne Block currently teaches academic literacies at the University of Malta, from where she also obtained an M.A. in Applied Linguistics and TESOL. Her research interests include multilingual language dynamics and professional learning. She has presented at multiple conferences and contributed to publications on these topics. She is also the co-editor of the IATEFL Research SIG newsletter. www.josianneblock.com
