Trainers and teachers can use AI to cut the time spent on first drafts: lesson outlines, practice questions, worked examples, rubrics and feedback comments. The rule that keeps it safe is simple. AI produces the draft, you supply the judgement. You decide what the learners need, you check every fact, and you adapt the material to the room in front of you.
Where AI saves the most preparation time
Preparation has a lot of repetitive structure, which is what AI handles well. The tasks below are where most trainers see the clearest gain.
- Lesson outlines. Give the topic, the audience, the time available and the outcome you want. Ask for a timed outline with activities, then reshape it.
- Examples and analogies. Ask for three different ways to explain one idea, pitched at a beginner, a manager and a specialist.
- Practice questions and quizzes. Request varied question types, with answer keys and a short reason for each answer.
- Case studies and role-plays. Generate realistic but fictional scenarios, such as a customer complaint or a project delay, for group discussion.
- Handouts and checklists. Turn your notes into a one-page summary or a step-by-step job aid.
- Feedback drafts. Turn your bullet-point observations into clear, kind written comments you then edit.
How to write a prompt that gives a usable lesson draft
Vague requests give generic material. A good prompt states who the learners are, what they already know, how long the session is and what they should be able to do afterwards. Add the format you want and any constraints, such as no jargon or Malaysian workplace examples.
For example, instead of "make a lesson on email writing", say who the learners are (new executives in a retail company), the length (a 90-minute session), the outcome (write a clear customer reply) and the format (a timed outline with two activities). Our guide to writing a good prompt for Claude covers the pattern in more detail.
Then iterate. Ask the AI to shorten the theory, add a group activity or make the examples more local. Treat the first answer as a starting point.
Check accuracy before anything reaches learners
AI can state wrong facts confidently, invent references and mix up details. In a classroom or a corporate workshop, a wrong statement from the trainer costs credibility. Build a short checking routine:
- Mark every fact, figure, date, name and rule in the draft.
- Verify each one against a source you trust, such as the official syllabus, a textbook or the product's own documentation.
- Remove anything you cannot verify, rather than keeping it because it sounds right.
- Test your own answer keys. Work through each quiz question yourself.
See how to check if AI output is accurate and what an AI hallucination is for the common failure patterns.
Keep the learner at the centre
Faster preparation is only useful if it buys better teaching. The time AI saves should go back into the parts only a person can do: noticing who is confused, adjusting the pace, asking a good follow-up question, and linking the topic to the learners' real work.
- Start from the outcome, not from what the AI produced. Delete any section that does not serve the outcome.
- Add your own experience. Stories and examples from your practice are what learners remember.
- Vary the activities. AI drafts can lean towards lectures and lists. Make room for practice and discussion.
- Be open with learners about how AI is used in your materials, and consider teaching them to use it well too.
Protect learner and organisation data
Do not paste names, marks, personal circumstances, student work with identifying details or confidential company material into an AI tool unless your institution or employer has approved that tool for that data. When asking for feedback drafts, describe the work generically or remove identifying details first. If you are unsure, ask your school, training manager or compliance team. Personal data raises PDPA questions; this is general information, not legal advice. Our article on putting customer data into Claude explains the thinking, and a team-wide AI usage policy sets the boundaries clearly.
Be careful with AI-generated feedback and marking
AI can help you phrase feedback, but it should not decide a grade or a performance outcome by itself. Learners deserve a human judgement they can question. A sensible split is that you assess the work, write your key points, and use AI to turn them into a clear, encouraging comment that you read before sending. Also remember that AI feedback can sound polished while missing what actually matters in the work.
A simple weekly workflow
- Write the learning outcome yourself in one or two sentences.
- Prompt for an outline and two activity options.
- Pick, rewrite and add your own examples.
- Generate practice questions, then verify answers and facts.
- Save prompts that worked in a shared library so colleagues can reuse them.
Where to learn this properly
Trainers who want a structured start can look at Agmo Academy, the training arm of Agmo Holdings Berhad. Agmo Studio is a Select Partner of the Anthropic Claude Partner Network. See AI training in Malaysia for options, or the Claude Certified Associate programme for a foundation-level path. HRD Corp claimability is subject to eligibility and HRD Corp approval, and prices are subject to SST.