I BELIEVE we are too quick to accuse artificial intelligence (AI) of damaging education.
Students use AI-driven chatbots like ChatGPT and we call them lazy. AI writes essays and we declare academic integrity under threat. AI demonstrates it can answer examination questions and our immediate response is to find better ways of detecting its use.
But there is a more uncomfortable question: If a machine can complete a university assignment well in 30 seconds, was that assignment really testing human intelligence in the first place?
Perhaps the real problem is not AI. Perhaps universities are still using forms of assessment designed for a pre-AI world. We continue to ask students to “describe”, “discuss”, “explain”, and “compare” material that machines can now process with ease. Then, when machines do exactly that, we blame the students. That is a strange response.
Universities cannot preserve 20th century assessment and then become angry when 21st century technology exposes its weaknesses.
AI is doing something useful to higher education. It is revealing which assignments were never really testing thought in the first place, but only the ability to gather, organise and reproduce information.
If we genuinely want to test human ability, change the question. Do not ask law students merely to state the law. Give them the law. Then give them conflicting facts, inconsistent authorities and incomplete information.
Ask them to decide. Ask them to defend the decision. Ask them to explain why the AI answer is wrong. Challenge them orally and require them to respond in real time. That is where the human contribution becomes visible.
In law, for example, a student should not merely be asked to identify the principle in a case. AI can do that. A more meaningful question is this: If two equally persuasive authorities point to opposite outcomes, which should prevail, and why?
That is not information retrieval. It is judgment. And higher education should be testing judgment.
This is why I am less persuaded by the current obsession with “detecting AI use”. We are asking the wrong question. The question is not: “How do we stop students from using AI?”
The better question is: “How do we design assessment that still has value even when students have AI?”
If the answer is to return to memory-heavy examinations, close laptops and prohibit technology, that is not innovation. It is retreat.
AI will not disappear after graduation. Lawyers will use it. Doctors will use it. Accountants, engineers, journalists, researchers, and administrators will use it.
An examination designed around a world without AI may therefore be testing competence for a world that no longer exists.
The challenge for universities is not to protect old assignments from AI. The challenge is to design tasks that AI cannot complete meaningfully without a human who genuinely understands the problem.
And perhaps this is the hardest point to admit: AI is not killing higher education. It is exposing the parts of higher education that may already have been intellectually obsolete, but which we have been reluctant to bury.
ASSOC PROF DR SHAHRUL MIZAN ISMAIL
Faculty of Law
Universiti Kebangsaan Malaysia
