01/12/2025

First page Previous page 10 Next page Last page

MONDAY | DEC 1, 2025

COMMENT by Zulkifli Musa

A CADEMIC publishing has long been a bastion of knowledge built upon rigorous peer review and intellectual integrity. Its purpose is to ensure that scholarly books and journals reflect high standards of research. Today, however, this trusted ecosystem faces a growing threat from the rapid rise of generative artificial intelligence (AI). Generative AI tools such as ChatGPT, Gemini and DeepSeek can instantly summarise research, produce essays and even draft full scientific articles. For students and researchers, these tools seem like convenient shortcuts that save time and boost productivity. Yet, within the world of academic publishing, the impact is far more complex. The integrity of scholarly communication is being quietly tested and the industry is struggling to keep pace. A decade ago, plagiarism primarily involved copying text from existing sources. Now, generative AI can produce original phrasing that does not technically copy anyone, yet may still contain false claims or misleading sources. AI hallucinations lead to fabricated data, incorrect information or fictitious research citations. Recent incidents have seen academics caught using fabricated references and manipulated findings. Once published, such work becomes a legitimate source for future papers, snowballing into a cycle of misinformation. This challenge has already reached Malaysian universities and publishers. Some institutions are experimenting with AI detection tools to filter out machine-written manuscripts. However, detection remains unreliable. AI content can be altered, paraphrased or edited by humans to bypass scrutiny. Relying solely on policing tools will not solve the issue. Academic publishers must reassess the ethics and workflows that underpin scholarship. University publishers, many of whom operate with limited resources, now face added pressure to protect quality while improving international visibility. Requiring AI scrutiny for every submission will further stretch resources. If editorial teams are overwhelmed by questionable manuscripts or fabricated citations, publication standards will fall and the credibility of Malaysian scholarship will be damaged. Across Asia, the risks are heightened. Academic publishing in the region has grown rapidly due to rising output from China, India and Southeast Asia. Many universities require publication in indexed journals for academic promotion. This publish-or-perish culture has already fostered paper mills, citation manipulation and the trading of authorship.

Growing risks AI poses to academic publishing

Generative AI is now the newest tool within this ecosystem, enabling researchers to produce manuscripts quickly without genuine expertise. The temptation to rely on AI-generated arguments or fabricated references becomes especially strong when publication plays a significant role in career advancement. For Malaysia, the risks are twofold. First, there is the danger of local researchers unintentionally polluting their work with AI generated fabrications, leading to retractions and reputational harm. Second, there is a deeper threat to students. If undergraduates and postgraduates begin using AI to write essays and literature reviews without critical engagement, we may be nurturing scholars who lose touch with the principles of evidence and original thinking. The line between a helpful tool and a crutch that fosters academic dependency is dangerously thin. The solution is not to prohibit AI but to use it responsibly. Generative AI has legitimate value in scholarly publishing. It can help with language editing, particularly for researchers who are not native English speakers. It can assist in summarising literature, improving clarity and supporting the analysis of large datasets. The risk lies in using AI to replace human reasoning. When interpretation, analysis and argument-building are outsourced to machines, the intellectual core of scholarship begins to erode. Transparency must, therefore, be central to academic publishing. Manuscripts generated partly or entirely by AI should not be treated as equivalent to human-authored work. Reviewers need to know when AI has shaped the research narrative. Editors should be informed if AI was used to draft sections of a manuscript or suggest references. Publishers have a responsibility to understand how research is produced because their role is to protect the value of intellectual labour. Education is equally vital. Instead of banning AI tools, institutions should teach students how to use them ethically. Avoiding AI is unrealistic. The real task is to cultivate academic honesty in a digital era.

If undergraduates and postgraduates begin using AI to write essays and literature reviews without critical engagement, we may be nurturing scholars who lose touch with the principles of evidence and original thinking. – BERNAMAPIC

Students must recognise that scholarship is not simply about producing text; it demands reasoning, originality and critical judgment. A generation that outsources thinking to machines will weaken the foundation of academic culture. Malaysia now has an opportunity to lead in this area. The country’s academic publishing community is relatively small, enabling more efficient policy coordination than in regions with thousands of journals. If Malaysian publishers, universities and researchers work together to establish clear guidelines now, they can avoid the confusion unfolding elsewhere. This is a chance to build a responsible AI framework before the situation becomes unmanageable. The future of scholarly publishing will be shaped by how society responds to this moment. Generative AI is not a villain but a disruptive force. It can elevate research quality

or undermine it entirely. It can broaden access to academic writing or flood the literature with convincing falsehoods. The decision lies not with machines but with editors, educators, researchers and policymakers who must balance innovation with integrity. What academic publishing should fear is not technology itself but the loss of trust. If scholarly knowledge becomes unreliable, the damage will spread through education, healthcare, policymaking and public discourse. Human intellect must remain at the heart of scholarship. AI can support the creation of knowledge but it cannot replace the pursuit of truth. Allowing machines to generate facts for us risks turning knowledge into a product rather than a meaningful discovery. Zulkifli Musa is the head of AI and Technology Unit, Universiti Sains Malaysia Press. Comments: letters@thesundaily.com

Education matters remain an important component of Malaysian society. Today the landscape is fast paced with new exciting developments in technology such as AI, coding, virtual reality, and cybersecurity, leading the way for changes in curriculum development. To showcase the latest and most updated content, we invite you join us in 2025, and be part of our Education Focus!

Malaysian Paper

Made with FlippingBook flipbook maker