OCTOBER 29 — Imagine a university student being accused of academic dishonesty for using artificial intelligence (AI) to improve an assignment. Now imagine the same assignment being marked by a lecturer who uses AI to generate feedback, suggest grades or evaluate the student’s arguments. The student is expected to declare every use of AI, but the lecturer may not be required to explain anything. Is this academic integrity, or are we practising double standards?

This is an uncomfortable question Malaysian universities must confront. We cannot continue treating AI as a threat when students use it and a symbol of innovation when lecturers do. If universities want students to behave ethically, lecturers must demonstrate the same responsibility.

Let us be clear. AI is not the enemy of education. Blind dependence on it is. A student who copies an AI-generated essay without thinking has learned very little. Equally, a lecturer who accepts AI-generated marks or comments without carefully reviewing them has neglected an important professional responsibility.

The solution is not to ban AI but to rethink how we teach, learn and assess. We can learn from universities abroad. The University of Sydney in Australia has introduced a two-lane assessment approach. Some assessments are supervised to establish what students can do independently, while others openly allow AI as a learning tool. This recognises a simple reality: graduates must learn both to think for themselves and to work responsibly with technology.

University College London takes another important position. While it provides guidance for responsible student use of AI, its guidance states that lecturers should not use generative AI to mark students’ work because of concerns about reliability and bias. Meanwhile, guidance at the University of Minnesota encourages transparency, human review and the option of human feedback when AI is involved.

These examples offer valuable lessons for Malaysia. First, universities should introduce an AI Accountability Receipt for both students and lecturers. Students would briefly explain how AI helped them, which suggestions they rejected and what intellectual contributions were genuinely theirs. Lecturers who use AI for feedback should similarly disclose its role and confirm that they personally reviewed the work. Academic honesty must travel in both directions.

Second, we should introduce an innovative assessment called “Put AI on Trial.” Instead of asking students to write another predictable essay, give them an AI-generated answer containing questionable claims, weak arguments and possibly fabricated references. Their task is to expose the mistakes, correct the reasoning and defend their decisions. The marks should reward the student’s judgment, not the machine’s fluency.

Third, every major AI-assisted assignment could include a short “Defend Your Thinking” conversation. Students explain their choices, challenge the AI’s suggestions and demonstrate what they have learned. This need not be an intimidating oral examination. Even a three-minute discussion can reveal thinking that a polished written submission may conceal.

Fourth, lecturers must remain responsible for the marks they award. AI may help prepare rubrics, organise feedback or suggest improvements where institutional rules permit, but it should never become an unquestioned judge of student achievement. Students also deserve privacy protection, meaningful explanations of their grades and a fair opportunity to challenge errors.

Finally, fairness means ensuring that students have reasonable access to approved AI tools and proper training. Otherwise, students who can afford expensive subscriptions may gain advantages over those who cannot. Universities should assess intellectual ability, not purchasing power.

Malaysia has an opportunity to develop graduates who are not merely competent users of AI but thoughtful critics of it. This requires lecturers to model the ethical conduct they expect from students. Perhaps the real question universities should ask is no longer, “Did you use AI?” but rather, “What did you contribute that AI could not?” And lecturers must be prepared to answer that question too. We cannot teach integrity through rules that apply only to students. If AI is to transform higher education, accountability must begin on both sides of the classroom.

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