Why Stronger Students Gain More from GenAI and How to Support the Rest

Writing is one of the most fundamental skills of an educated person. It is not only about the ability to produce words on paper, but also about the skill of putting ideas into text, organizing them logically, and addressing the text to a specific audience. Fluent and coherent writing is a skill that demands knowledge, training, and constant practice. With the advancements in technology, more digital tools have appeared to assist individuals in writing. Generative Artificial Intelligence (GenAI) is stretching a helping hand to a writer, suggesting all sorts of possible assistance and even substitution in the process of writing. How do educators make sure that students gain the necessary writing skills and learn to use appropriate tools for strengthening and developing their skills?

Author: Olesia Kullberg

Transferring one’s ideas into written text is a challenging process of production that includes several stages. According to Johansson (2025), seminal cognitive theories define three main stages that the writer goes through: planning – deciding on the goal, topic, content, and text organization; translation – putting ideas into written words, phrases, and sentences; revision – assessing how the text meets the initial goals and possibly editing. The process is not always linear, and one can go through multiple cycles while working on the text. (Johansson 2025) Depending on the skills of the writer, the process might feel extremely chaotic and unproductive.

Writing skills of university students vary significantly, as do their generic study skills. The accessibility of GenAI’s tools and applications provides seemingly effective and necessary support to all students in their studies. However, the primary skills of students have a significant impact on the way they use GenAI in their studies, threatening to increase the gap in skills even further.

Motivated and skilled students benefit from GenAI the most

Regardless of the existing concern about the deteriorating impact of GenAI on the quality of written texts, current research shows that the use of GenAI in teaching and studying is beneficial for improving students’ writing skills. Through semi-structured university teacher interviews, Marzuki et al. (2023) revealed that teachers perceived a notable positive impact in their students’ written work completed during classroom lessons with the use of GenAI tools. In particular, the improvement was identified in idea development, clarity of content, and the organization and coherence of students’ texts. Zhao (2025) highlights that there is clear evidence of improvements in student’s writing skills in language precision, content summarization, and creative facilitation due to the tailored support that AI tools can provide students with. However, both studies insist on the importance of correct pedagogical integration of GenAI tools into the study process that includes guidance, supervision and constant reflection. When used as part of a structured and thoroughly organized study process, GenAI tools help students learn how to write better texts.

On the other hand, the study by Cui et al. (2025) shows that concerns about the negative impact of GenAI on writing skills are not rootless. The question is about how different students use GenAI tools and what outcomes they achieve. The researchers identified four GenAI-use patterns among students while completing writing assignments. The first group, who benefitted from GenAI tools the most, were active collaborators – students with a high level of autonomy, motivation, and self-regulation, who used AI as a dialogue partner to refine ideas they already had. The second group, selective users with moderate autonomy and motivation and less consistent self-regulation, consulted AI mainly for revision, benefitting mostly from proofreading. The third group were identified as passive adopters; due to lower autonomy, weaker self-regulation, and externally driven motivation, they relied heavily on AI-generated content with minimal reflection or evaluation of the outcomes. Finally, the fourth group were the minimal users, who had low familiarity with AI in general, which resulted in their hesitant and ineffective use of AI in writing. The study demonstrates that students use GenAI tools in markedly different ways, ranging from active collaboration to passive reliance. (Cui et al. 2025)

Educators highlight the idea of using GenAI as a collaborative partner, not as a substitute for completing tasks. In fact, the real power of GenAI in studies lies in its ability to support a human in the process of completing tasks. Nevertheless, getting to the level of active collaboration with GenAI is a long process that demands new approaches and study content from the teachers. Most students are not ready to collaborate with GenAI ethically and effectively without teacher support.

Supporting the development of students’ writing skills

As the writing process consists of three major stages – planning, translation, and revision – students should be clearly instructed on self-regulation processes that impact their ethical and effective use of GenAI at each stage. For example, defining the goal of the text and the main take-away the reader should get from reading the text is crucial to complete on their own, as this will allow the writer to stay on the same track while working on the text. GenAI tools help in creating more effective text when they are used strategically at the translation and revision stages. In practice, strategic use of GenAI while writing means going through a conversation, asking for several suggestions, evaluating how the provided suggestions align with the original goal of the text, and writing the text in the writer’s own words and style. Assignment instructions should include clear guidance on the possible use of GenAI tools on each stage of the writing process.

Instructions alone are not yet sufficient for making students collaborate with GenAI effectively while writing. Lack of motivation and low self-regulation can partly be caused by the challenging nature of writing assignments – insufficient skills can be one reason for using GenAI as a substitute while writing. The solution could be to implement practice and testing writing assignments. Practice assignments can help students try different tools that can assist them in creating their own texts. These assignments can include many steps and require longer text creation. While testing assignments can be a short review of the same text written by the student without any tools. Testing assignments can be arranged in the form of a written exam under supervision. They would allow students to demonstrate the skills they have acquired working on the practice assignments. Evaluation could focus on the process including both practice and testing assignments, or only on the final version produced by the student without any tools. A combination of practice and testing assignments could release the pressure from the idea of writing a perfect text entirely on their own. At the same time, knowing that there is help from GenAI that one can use following instructions and that there will be a testing point at the end, could motivate students to proceed with suggested method.

Provided ideas can be relatively easy to implement in courses with regular classroom lessons. However, online courses mostly lack the possibility of inviting students to sit a supervised exam. Thus, guidance and support in the process of writing play a paramount role in online courses. After all, it is not the fear of a supervised exam, but the understanding of the importance of the ability to express one’s own ideas clearly and effectively in writing that should be a driving force for students in their journey of revealing the potential of the new possibilities that we all have in the form of GenAI tools.

GenAI tools provide support and improve the quality of writing, but depending on the skill set of a student, the benefit from those tools varies greatly. As educators, our role is to support the development of one of the crucial skills, i.e. writing, and this is done through guidance, assignment design and implementation, and supervised writing tasks.

References

Cui, Y. He, W. Du, X. Zeng, M. & Liu, D. 2025. The Impact of AI Writing Assistants on Academic Writing Performance: From the Perspective of Subjectivity. International Journal of Distance Education Technologies. Vol. 23 (1), 1–28. Cited Jan 27 2026. Available at https://doi.org/10.4018/IJDET.391326

Johansson, V. 2025. Writing processes: What happens when a text is born, and how may (AI) writing tools shape future writing processes? Kristianstad University. Cited Jan 26 2026. Available at https://researchportal.hkr.se/en/publications/writing-processes-what-happens-when-a-text-is-born-and-how-may-ai/

Marzuki, Widiati, U. Rusdin, D. Darwin, & Indrawati, I. 2023. The impact of AI writing tools on the content and organization of students’ writing: EFL teachers’ perspective. Cogent Education. Vol. 10 (2). Cited Jan 27 2026. Available at https://doi.org/10.1080/2331186X.2023.2236469

Zhao, D. 2025. The impact of AI-enhanced natural language processing tools on writing proficiency: an analysis of language precision, content summarization, and creative writing facilitation. Education and Information Technologies. Vol. 30, 8055–8086. Cited Jan 27 2026. Available at https://doi-org.ezproxy.saimia.fi/10.1007/s10639-024-13145-5

Author

Olesia Kullberg is a Senior Lecturer at Language Centre of LAB University of Applied Sciences. Her expertise lies in professional communication and digital pedagogical solutions.

Illustration: https://pxhere.com/en/photo/876111 (CC0)