JULY 13 — Walk into any university campus in Malaysia today and you will witness a profound structural shift in how students work. The abundant presence of Generative AI tools like ChatGPT, Gemini and many more is widely celebrated as a democratizing force for education and a massive boost to productivity.

Nonetheless, the ability to generate a 3,000-word essay in seconds does not equate to the true actual learning process.

Underneath the surface of this technological convenience lies a growing and hidden cost: the erosion of our students’ critical thinking and problem-solving capacities. In our rush to embrace educational efficiency, we are inadvertently encouraging our youth to outsource their intellectual labor to machines.

Historically, the act of writing and researching in higher education was not just about producing a final product but it was the very mechanism of thinking. Cognitive science dictates that true learning requires “desirable difficulties.” The friction of wrestling with complex ideas, organizing a messy argument and occasionally failing is what indeed builds intellectual resilience and cements neural pathways.

When a student outsources a complex assignment to an AI, they completely bypass this vital critical struggle. They in fact transition from being the “author” of an original thought; to a mere “editor” of algorithmically generated content. The AI absorbs the heavy cognitive load while leaving the student intellectually impoverished when separated from their smartphone or laptop.

This phenomenon known as “cognitive offloading” is uniquely challenging for a developing nation like Malaysia.

Generative AI models are predominantly trained on vast datasets from the Global North, heavily skewing towards Western cultural, legal and economic norms. When local students unthinkingly rely on these tools to solve localized problems such as drafting a sustainability strategy for a local Small and Medium Enterprises (SME); the AI often “hallucinates” or attempts to map Western frameworks onto contexts that simply do not apply in the country.

If we do not actively cultivate critical AI literacy, we risk an “algorithmic colonization” of our educational system. Additionally, future leaders may internalize a homogenized, Western-centric worldview that overlooks local socio-economic realities, the nuances of Islamic finance or the specific needs of local diverse communities.

This professional threat is particularly severe in fields like business, management, corporate governance and alike. For instance, corporate governance is not a discipline of absolute mathematical truths but requires situational judgment, stakeholder empathy and ethical reasoning, notably in gray areas.

If graduates enter the workforce only knowing how to simulate compliance rather than strategically navigate ethical dilemmas, they will struggle to manage the real-world complexities of local supply chains, labor dynamics and institutional accountability.

Eventually, this might place the long-term resilience of Malaysia’s corporate sector at risk.

This vulnerability extends across the broader economic landscape. As AI inevitably automates routine administrative workflows, the premium on human labor will shift entirely toward unstructured problem-solving and strategic foresight. If our graduates are trained solely to act as passive editors of machine-generated content, they risk becoming easily replaceable by the next software update. To remain competitive, Malaysia requires a workforce capable of managing AI, not a workforce rendered obsolete by it.

So, what is the solution?

Banning AI or relying entirely on flawed plagiarism detection software is a futile cat-and-mouse game. Generative AI is rapidly becoming an unavoidable embedded tool in the modern macroeconomic workplace.

Instead, Malaysian higher education institutions and policymakers must urgently pivot toward a framework of “cognitive integrity.” We need to shift our pedagogical focus from merely assessing the final product to thoroughly evaluating the intellectual process.

This means a return to iterative drafting, where students must gradually track and defend the evolution of their thoughts. It means maintaining or reintroducing (where need be) the scaled-down oral defenses (viva voce) for undergraduate coursework, forcing students to defend their ideas and methodology in real-time under pressure. This is amongst assessment mechanisms that an AI simply cannot replicate for them. Additionally, improvised Problem-Based and Project-Based Assessments and Activities or PBAs in class activities; could also ensure less reliance on students in getting quick answers from AI.

Likewise, students must be taught that AI is not an all-knowing oracle but merely a predictive text engine. Curriculums must evolve to position AI as an intellectual sparring partner rather than a shortcut. Students should be actively tasked with critiquing AI-generated arguments, identifying their inherent biases and correcting their logical flaws.

Systemically, accreditation frameworks must be recalibrated to emphasize these critical competencies, ensuring that technological adoption does not outpace intellectual consistency.

The goal of Malaysia’s higher education system, particularly in the spirit of emphasis on innovation and critical thinking should not merely be to produce graduates who can rapidly generate content. In an era where information is infinitely and cheaply generated by machines, human economic value will lie entirely in the ability to critically analyze and synthesize that data.

Therefore, the relevant parties should teach students on not only how to prompt the machine; but how to out-think it. The future intellectual capital of our nation depends on it.

* Hasroleffendy Hassan is a senior lecturer in the Faculty of Business and Management, Universiti Teknologi MARA, Kedah Branch.

** This is the personal opinion of the writer or publication and does not necessarily represent the views of Malay Mail.