INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XII December 2025  
Teaching in theAge ofAI: Lecturer Perceptions ofAI-Assisted Student  
Writing in a Private University in Perak, Malaysia  
Amanpreet Kaur Gurdarshan Singh1,Sshivapriya Sundran2,Ruban Paul Durai3,Pavinder Kaur Girn  
Baldev Singh4,Noor Eleena Nordin5  
Faculty of Social Sciences, Quest International University, Malaysia  
*Corresponding Author  
DOI: https://doi.org/10.47772/IJRISS.2025.91200031  
Received: 27 November 2025; Accepted: 05 December 2025; Published: 31 December 2025  
ABSTRACT  
The rapid integration of artificial intelligence (AI) writing tools in higher education has generated growing  
debate about their influence on student learning, writing development, and academic integrity. This study  
explores lecturers’ perceptions of AI-assisted student writing at a private university in Perak, Malaysia. Semi-  
structured interviews were conducted with five lecturers to examine how AI tools such as ChatGPT are  
shaping students’ writing practices, the extent to which students rely on these technologies, and the challenges  
and opportunities they pose. Findings indicate that AI tools have improved students’ grammar, organization,  
and vocabulary, especially among those who previously struggled with language proficiency or structuring  
ideas. However, lecturers also observe increasing dependency on AI-generated content, insufficient  
understanding of the submitted work, and inconsistent citation practices. These concerns highlight broader  
questions about authentic learning, ethical use, and the future of academic writing instruction. The study offers  
insights into how universities can balance AI’s benefits with the need to maintain academic integrity and foster  
independent thinking, and proposes ways to integrate AI responsibly into the learning environment.  
INTRODUCTION  
Academic writing remains a central component of higher education, serving as a foundation for critical  
thinking, research engagement, and scholarly communication. Academic writing courses aim to develop  
students’ abilities in structuring arguments, synthesizing information, demonstrating originality, and adhering  
to academic integrity principles. They also prioritise skills such as critical analysis, logical organization, and  
accurate referencing (Teng & Wang, 2023; Cheong et al., 2023; Mendoza et al., 2022; Schillings et al., 2023).  
With the increasing presence of digital tools, the landscape of academic writing has experienced substantial  
transformation.  
The emergence of AI-powered writing tools has intensified questions about their role in shaping students’  
writing abilities. Although tools such as Grammarly and Turnitin support proofreading and similarity  
detection, more advanced systems like ChatGPT can generate entire paragraphs, provide explanations, refine  
sentence structures, and offer stylistic suggestions. These capabilities raise concerns about the extent to which  
AI may influence students’ creativity, critical thinking, and engagement in the writing process (Teng et al.,  
2022; Zhao et al., 2023). Scholars also highlight that AI-generated writing often lacks depth, analytical  
coherence, and nuanced argumentation elements that require human reasoning and conceptual understanding  
(Schillings et al., 2023).  
Educators increasingly express difficulty in distinguishing student-authored content from AI-generated text  
(Grimes et al., 2023). Although AI detection tools have been introduced, their accuracy remains questionable,  
as they often misidentify human writing as AI-generated or fail to detect substantial AI involvement (Warner,  
2023; Elkhatat et al., 2023). Benarab (2024) emphasizes that no single detection method is reliable, calling for  
a combination of instructor judgment, process-based assessments, and technological literacy.  
Beyond detection issues, there remains ongoing debate about the pedagogical implications of AI tools. While  
they can support language development and reduce writing anxiety, they may also compromise academic  
Page 348  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XII December 2025  
integrity and hinder the development of metacognitive and independent writing skills (Perkins, 2023). This  
study aims to understand how lecturers perceive the influence of AI tools on student writing performance,  
learning behaviour, and ethical engagement.  
The guiding research question is:  
How do lecturers perceive the influence of AI writing tools on students’ writing development and academic  
practices in a private university in Perak, Malaysia?  
LITERATURE REVIEW  
Theoretical Framework  
This study adopts the Technology-Based Learning Model proposed by Hsu et al. (2012), which outlines how  
technology integration can be examined through three components: application domains, research methods,  
and research issues. This model has been widely applied in synthesising research on emerging technologies in  
education, including studies involving chatbots and AI-driven learning tools (Hwang & Chang, 2023; Liu &  
Hwang, 2023).  
Hwang and Chang (2023) found that chatbots have been most widely used in language education, followed by  
disciplines such as engineering and computer science. Most studies employ quantitative methods, though  
qualitative and mixed-methods designs are increasingly used to explore learner perceptions and challenges in  
technology-enhanced environments. Key research gaps include limited understanding of effective learning  
designs and the lack of standardised frameworks for integrating AI tools into instruction.  
Building on this model, the present study examines lecturers’ perceptions of AI writing tools, focusing on how  
these tools affect students’ writing practices, cognitive engagement, and academic behaviour. Empirical studies  
such as those by Bin-Hady et al. (2023), Mizumoto and Eguchi (2023), Mohamed (2024), Yan (2024), and  
Young and Shishido (2023) provide insights into how AI tools influence language learning, writing support,  
and learner autonomy. These works suggest that while AI provides scaffolding, its impact largely depends on  
students’ digital literacy, motivation, and ethical awareness.  
METHODS  
This research adopts a qualitative design to explore lecturers’ perceptions of AI writing tools within the higher  
education context. Semi-structured interviews were conducted with five lecturers from a private university in  
Ipoh, selected through purposive sampling due to their experience supervising and assessing assignments in  
which AI tools were used.  
Each interview lasted 30 to 45 minutes and was conducted either in person. The interview protocol focused on  
key areas: changes observed in student writing quality, the extent of student reliance on AI-generated content,  
the benefits and challenges of AI-assisted writing, and concerns regarding academic integrity. All interviews  
were audio-recorded with consent and transcribed verbatim.  
Thematic analysis was applied to identify recurring patterns and divergent viewpoints across the interviews.  
This approach provided an in-depth understanding of how AI tools are viewed by educators and how these  
perceptions inform teaching practices and assessment decisions.  
FINDINGS AND DISCUSSION  
The interview data revealed a complex mixture of appreciation and concern among the five lecturers regarding  
the role of AI in students’ academic writing. All lecturers agreed that AI tools have contributed positively to  
the linguistic and structural aspects of students’ work. Students who previously struggled with grammar,  
vocabulary, and idea organisation demonstrated noticeable improvement after incorporating AI assistance.  
This was especially evident among weaker writers, who became more capable of producing organised and  
coherent assignments. One lecturer shared that even students with strong language proficiency benefitted from  
AI, particularly when refining the logical flow of their arguments and structuring their ideas more effectively.  
Page 349  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XII December 2025  
Lecturer 1 highlighted that students’ writing had improved markedly, especially in terms of grammar,  
structure, and vocabulary. The lecturer perceived ChatGPT as a supportive tool that could enhance students’  
critical thinking by offering ideas and helping them expand their perspectives. Lecturer 4 reinforced this  
observation by noting that assignments were being submitted with fewer grammatical errors and in a more  
timely manner. Lecturer 5 similarly pointed out that students who had previously been considered problematic  
were now able to produce more organised work, suggesting that AI offers meaningful scaffolding for those  
who need additional support. These positive perceptions align with recent literature indicating that AI can  
enhance students’ linguistic accuracy and cognitive engagement when used as a learning partner rather than a  
shortcut.  
Despite these benefits, the lecturers expressed strong concern about students’ increasing reliance on AI without  
genuine engagement in the learning process. Lecturer 2 observed that many students submitted assignments  
generated almost entirely by ChatGPT, often without understanding the content. This became clear when  
students were asked to explain their work or reproduce similar content independently. Lecturer 3 echoed this  
concern, noting that some students contributed very little of their own thinking and depended heavily on AI to  
construct their arguments. Such behaviours mirror findings from current research that highlights the risk of  
superficial learning when students outsource cognitive tasks to generative tools.  
Ethical issues also emerged as a major theme across the interviews. Lecturers reported that students frequently  
copied AI-generated content directly without acknowledging the source. The absence of proper citation  
practices was noted by Lecturer 3, while Lecturer 2 described cases where entire responses were lifted from AI  
without modification. Although Lecturer 1 believed plagiarism could be controlled through clear citation  
requirements, other lecturers felt that students did not fully adhere to these expectations. These concerns  
resonate with wider academic discussions about the challenges of maintaining academic integrity in an AI-  
mediated learning environment, particularly given the limitations of AI detection tools and inconsistencies in  
students’ digital literacy.  
Another issue highlighted by the lecturers was the diminishing involvement of students in the writing process.  
Several lecturers observed that students often treated AI as a replacement for personal effort rather than as an  
academic resource meant to support learning. Lecturer 2 emphasised that students appeared less stressed  
because the AI completed much of the cognitive work for them, yet this lack of active participation  
undermined the experiential nature of writing. Lecturer 3 similarly expressed concern that student dependency  
on AI weakened their ability to develop critical thinking, independent argumentation, and sustained writing  
practices. This is consistent with existing scholarship warning that while AI can scaffold learning, it may also  
inadvertently reduce students’ metacognitive engagement if used uncritically.  
Despite the risks of over-reliance, the lecturers acknowledged that AI has reduced writing anxiety, especially  
for students who lack confidence in their linguistic abilities. Several lecturers reported that assignments were  
being submitted more consistently, possibly because AI helped students overcome challenges related to idea  
generation and initial drafting. This observation is supported by emerging research suggesting that AI, when  
used effectively, can act as a confidence-building tool that enables students to complete tasks they might  
otherwise avoid.  
Collectively, the lecturers’ perspectives illustrate the dual nature of AI-assisted writing. On one hand, AI  
improves accuracy, coherence, and organisation, benefitting both strong and weak writers. On the other hand,  
misuse of AI leads to superficial learning, ethical breaches, and reduced student engagement. These findings  
suggest that while AI holds significant pedagogical potential, it must be integrated within a structured  
framework that promotes critical use, ethical awareness, and sustained student participation. Such balance  
aligns with scholarly recommendations advocating for AI literacy, explicit guidelines, and thoughtful  
curriculum design to ensure that technology enhances rather than replaces meaningful learning.  
Limitations  
This study has several limitations. AI technologies evolve rapidly, and the capabilities of writing tools and  
detection systems may change significantly over time. Current literature and lecturer perceptions may therefore  
become outdated quickly. The study is based on a small sample from a single private university, which limits  
Page 350  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XII December 2025  
the generalisability of the findings to broader educational contexts. Experiences in multilingual, public, or  
international institutions may differ, particularly where writing conventions or levels of digital literacy vary.  
Moreover, the study relies on self-reported experiences through interviews, which may not fully capture the  
complexities of students’ actual writing behaviour or the extent of AI usage. Longitudinal research would  
provide deeper insight into how sustained AI use influences writing development, academic identity, and  
metacognitive strategies.  
RECOMMENDATIONS  
Based on the findings, several recommendations are proposed to improve the integration of AI tools in  
academic writing instruction while safeguarding academic integrity:  
1. Develop clear institutional guidelines  
Universities should create transparent policies that clarify acceptable and unacceptable uses of AI writing  
tools. Instead of blanket bans, guidelines should emphasise ethical use, attribution, and critical engagement.  
2. Integrate AI literacy into writing instruction  
Educators should explicitly teach students how to use AI tools such as paraphrasers, editors, and feedback  
systems responsibly and effectively. This includes training on prompt creation, revision strategies, and  
evaluating AI-generated content.  
3. Shift assessment design toward higher-order skills  
Because AI can generate basic essays easily, assessments should prioritise critical thinking, reflection,  
personalised analysis, and process-based tasks (drafts, logs, peer reviews) that reduce dependence on  
automated outputs.  
4. Use detection tools cautiously  
Given documented detection inaccuracies (Elkhatat et al., 2023; Warner, 2023), AI-detectors should never be  
used as the sole evidence for academic misconduct. Instead, they should serve only as preliminary indicators  
supported by human judgment.  
5. Support students’ writing self-efficacy  
Instructors should incorporate metacognitive strategy instruction (Ishak et al., 2025) and confidence-building  
tasks to ensure that students continue to develop genuine writing competence rather than relying solely on  
automation.  
6. Provide professional development for educators  
Lecturers need ongoing training to understand AI capabilities, limitations, and pedagogical applications so  
they can guide students more effectively.  
CONCLUSION  
AI-assisted writing tools have introduced new dynamics into academic writing practices in higher education.  
While they offer substantial benefits by improving language accuracy, enhancing clarity, and supporting  
students who struggle with writing, they also pose challenges related to ethical use, dependency, and authentic  
learning. Lecturer experiences in this study reveal that AI can be a valuable tool for scaffolding student  
learning, but its effectiveness depends on students’ willingness to critically engage with the writing process  
and understand the content they produce.  
The findings reinforce the need for clear institutional guidelines, AI literacy instruction, and assessment  
designs that promote higher-order thinking and originality. Universities must strike a balance between  
Page 351  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XII December 2025  
embracing technological innovation and preserving the integrity of academic writing. AI should serve as a  
supplementary aid rather than a substitute for students’ reasoning, creativity, and intellectual growth. With  
proper guidance and ethical frameworks, AI tools can enrich the learning experience while maintaining the  
core values of academic scholarship.  
ACKNOWLEDGEMENT  
This publication is supported by Quest International University under the QIU Academic Publication Funding.  
REFERENCES  
1. Aljuaid, H. (2024). The impact of artificial intelligence tools on academic writing instruction in higher  
education: A systematic review. Arab World English Journal (AWEJ) Special Issue on ChatGPT, April  
2024. Available at SSRN: https://ssrn.com/abstract=4814342  
2. Benarab, I. H. (2024). Detection of AI-generated writing in students’ assignments: A comparative  
analysis of some tools’ reliability. Atras Journal, 5(Special Issue), 271–286.  
3. Elkhatat, A. M., Elsaid, K., & Almeer, S. (2023). Evaluating the efficacy of AI content detection tools in  
differentiating between human and AI-generated text. International Journal for Educational Integrity,  
19(1), 1–16. https://doi.org/10.1007/s40979-023-00140-5  
4. Grimes, M., Von Krogh, G., Feuerriegel, S., Rink, F., & Gruber, M. (2023). From scarcity to abundance:  
Scholars and scholarship in an age of generative artificial intelligence. Academy of Management  
Journal, 66(6), 1617–1624.  
5. Ishak, Maziahtusima & Oderinde, Idayat & Ahmad, Sakinah. (2025). The Role of Metacognitive  
Strategies in Enhancing Learning Outcomes and Educational Efficiency: A Systematic Review of  
Quantitative, Qualitative and Mixed-Method Studies. International Journal of Academic Research in  
Business and Social Sciences. 15. 10.6007/IJARBSS/v15-i4/24964.  
6. Perkins, M. (2023). Academic integrity considerations of AI large language models in the post-pandemic  
era: ChatGPT and beyond. Journal of University Teaching & Learning Practice, 20(2), 3–16.  
https://doi.org/10.53761/1.20.02.07  
7. Schillings, M., Roebertsen, H., Savelberg, H., & Dolmans, D. (2023). A review of educational dialogue  
strategies to improve academic writing skills. Active Learning in Higher Education, 24(2), 95–108.  
https://doi.org/10.1177/1469787418810663  
8. Teng, M. F., & Wang, C. (2023). Assessing academic writing self-efficacy belief and writing  
performance in a foreign language context. Foreign Language Annals, 56(1), 144–169.  
https://doi.org/10.1111/flan.12638  
9. Teng, M. F., Qin, C., & Wang, C. (2022). Validation of metacognitive academic writing strategies and  
the predictive effects on academic writing performance in a foreign language context. Metacognition and  
Learning, 17(1), 167–190.  
10. Warner, A. (2023). The false promise of generative AI detectors. Multilingual. Available at  
https://multilingual.com/the-false-promise-of-generative-ai-detectors/  
11. Zhao, X., Sbaffi, L., & Cox, A. (2023). The digitization of writing in higher education: Exploring the use  
of Wordtune as an AI writing assistant. https://doi.org/10.31219/osf.io/uzwy7  
Page 352  
www.rsisinternational.org