SEPT 12 — Every semester, academics face an increasingly familiar reality: students are using artificial intelligence (AI) tools such as ChatGPT and Gemini as part of their academic work.
AI is no longer something universities can simply keep outside the classroom; it is embedded in the digital environment where undergraduate assignments, postgraduate research and everyday learning now take place.
Yet much of the institutional response has focused on control — oral defences, offline in-class writing, restrictions on AI use and controversial AI-detection tools.
While these measures have a place, relying on them as our primary defence risks missing a much bigger question: what exactly are we trying to assess?
A persistent flaw in our current thinking is the assumption that students can somehow be separated from technology.
The generation entering our lecture halls has grown up with smartphones and instant access to information.
Technology is not an occasional addition to their learning environment; it is part of it.
Returning entirely to analogue forms of assessment may solve one problem while creating another, especially as students enter workplaces where the ability to work critically and responsibly with AI is increasingly valuable.
The real challenge is not that students are using software to assist with academic tasks.
It is that many assessment models still place enormous weight on the final product — the polished report, essay or written answer — while revealing relatively little about the human process behind it.
If an assignment can be completed convincingly with a single AI prompt, perhaps we should also question what that assignment was designed to measure.
Protecting academic integrity requires more than catching students using AI; it requires us to reconsider what evidence of learning should look like.
This responsibility does not belong to students or lecturers alone; it must be shared across the entire higher education ecosystem.
In practice, rather than asking students to submit every conversation they have had with an AI tool, assessments could require a concise AI-use log or selected prompt trail.
Students could identify key prompts used, highlight suggestions they accepted or rejected, point out factual errors and explain how they verified information using credible sources.
The purpose is to make the student’s intellectual contribution visible: what they questioned, changed, verified and ultimately decided.
Assessment could also require students to interrogate AI-generated answers rather than simply reproduce them.
They might compare different AI responses, test an AI-generated argument against academic literature or explain why a seemingly convincing answer is actually weak.
This shifts the student’s role from passive recipient to critical evaluator.
Furthermore, universities could rethink the dominance of the one-off final submission.
Instead of receiving only a polished essay at the end of the semester, lecturers could assess selected stages of the student’s intellectual journey, from the initial idea and AI-assisted exploration to verification and final revision.
None of this means academic misconduct should become acceptable because AI exists.
Passing off AI-generated work as one’s own or fabricating references remain serious integrity concerns.
But academic integrity should not become weaker in the age of AI.
It should become more demanding.
Students should be expected not merely to produce an answer, but to demonstrate judgment, verification, intellectual ownership and the ability to defend the choices behind it.
The goal of modern higher education must extend beyond training students to produce polished documents.
AI can generate remarkably fluent text in seconds.
But students cannot outsource the responsibility to judge whether that text is accurate, credible, ethical and worth accepting.
Perhaps the question we should ask is no longer simply, “Did the student use AI?”
A more meaningful question is, “What intellectual work did the student still have to do?”
If universities can redesign assessment around that question, AI will not signal the end of academic integrity.
It may instead force us to define it more clearly than we ever have before.
* Dr Nor Farah Hanis Zainun is a senior lecturer at the School of Business Management, Universiti Utara Malaysia.
* This is the personal opinion of the writer or publication and does not necessarily represent the views of Malay Mail.
