Answer-first summary
AI can help lecturers give faster, more consistent feedback and design better assessments, but it should support human judgement, not replace it. This guide shows Malaysian lecturers how to use generative AI to draft rubrics, generate practice questions, speed up formative feedback, and redesign assessments so they remain meaningful in an AI era, while keeping fairness and academic integrity intact.
Where AI genuinely helps with assessment
Marking and feedback are among the most time-consuming parts of teaching. AI can ease that load in specific, well-bounded ways, freeing lecturers to focus on the higher-value work of mentoring and judgement.
The key principle is simple: use AI to draft, then refine with your own expertise. AI should produce a starting point that you check, correct, and personalise. It should never issue a final grade or unreviewed comment to a student.
Four practical uses for lecturers
1. Drafting rubrics faster
Describe your assignment and learning outcomes to an AI tool and ask it to propose a rubric with criteria and performance levels. You then edit the wording, adjust the weightings, and align it to your course standards. This turns an hour of rubric writing into fifteen minutes of refinement.
2. Generating practice questions and quizzes
AI can quickly produce a bank of practice questions at different difficulty levels from your lecture notes or a topic outline. Review each item for accuracy, remove anything misleading, and you have low-stakes practice material that helps students self-assess before formal exams.
3. Speeding up formative feedback
For drafts and formative work, AI can help you generate first-pass comments tied to your rubric. You then edit those comments so they reflect your voice, the specific student’s work, and your professional standards. The goal is faster feedback, not generic feedback, so always personalise before sending.
4. Reviewing your own assessment design
Paste an existing assignment into an AI tool and ask whether it can be completed by AI alone. If the answer is yes, that is a signal to redesign the task so it measures genuine student thinking.
Redesigning assessments for an AI era
If an assignment can be answered well by a chatbot in seconds, it no longer measures what you intend. The solution is not to ban AI but to design assessments that reward the things AI cannot do on a student’s behalf.
Strong AI-resilient assessment designs include: oral defences where students explain their reasoning, process portfolios that show drafts and decisions over time, locally grounded tasks tied to Malaysian contexts and the student’s own community, reflective writing on the student’s learning journey, and assignments that explicitly require students to critique or improve AI-generated output. These approaches shift the focus from the final product to the thinking behind it.
This kind of assessment redesign is a core part of practical AI in education training, because it requires both an understanding of the tools and pedagogical judgement about what each assessment should measure.
Keeping feedback fair and accurate
AI-generated feedback carries real risks if used carelessly. AI can be confidently wrong, can carry bias, and can produce comments that sound polished but miss the substance of a student’s work. Three safeguards keep feedback fair:
First, verify every factual claim and citation the AI produces before passing it to a student. Second, never let AI assign a final grade; grades require human accountability. Third, treat student work as confidential and avoid pasting identifiable personal data into public AI tools, in line with privacy expectations.
A simple workflow lecturers can adopt this semester
- Pick one assessment. Start small with a single assignment rather than redesigning your whole course.
- Use AI to draft the rubric, then refine it yourself.
- Generate a practice question bank and check each item.
- For formative drafts, use AI to speed up first-pass comments, then personalise.
- Reserve all final grades and integrity decisions for yourself.
- Ask students for feedback at the end of term and adjust.
This staged approach lets you capture the time savings of AI while protecting the quality and fairness that students depend on.
Frequently asked questions
Can AI grade student work for me?
AI can help draft feedback and suggest scores against a rubric, but it should not assign final grades. Grading requires human accountability, contextual judgement, and fairness that AI cannot guarantee. Use AI as a drafting assistant, not a marker.
Is it ethical to use AI to give students feedback?
Yes, provided you review and personalise every comment, verify accuracy, and protect student privacy. Many institutions consider AI-assisted feedback acceptable when the lecturer remains fully responsible for the final feedback delivered.
How do I stop students from using AI to complete assessments?
Rather than trying to block AI entirely, redesign assessments to reward reasoning, process, and locally grounded thinking. Oral defences, draft portfolios, and reflective tasks are far harder to outsource to a chatbot.
Where can lecturers learn these skills?
Dr Muhamad Hariz, Senior Lecturer at UPSI and HRD Corp Accredited Trainer, delivers workshops for lecturers on AI-assisted assessment and feedback. See his classroom approach in How I Use AI in My University Courses and his guidance on classroom rules in A Practical AI Policy for Classrooms in Malaysia.
Work with Dr Hariz
If your faculty wants hands-on training in AI-assisted assessment and feedback, reach out through the contact page to arrange a workshop or advisory session for your lecturers.
Written by Dr Muhamad Hariz Adnan — Senior Lecturer at UPSI, PhD in IT (Universiti Teknologi PETRONAS), and HRD Corp Accredited Trainer specialising in AI and digital transformation in education.