The parts only you should write are the ones that carry your lived experience: your personal stories, the frameworks you developed through trial and error, your honest opinion on what actually works versus what just sounds good, and any moment where a student needs to feel your direct voice coaching them through something hard. AI can draft supporting content, but it cannot replace what you have personally learned.
The Line Between Voice and Structure
AI is very good at structure. It can scaffold a curriculum, write lesson introductions, generate examples, produce worksheets, and summarise complex ideas clearly. What it cannot do is draw on the specific Tuesday afternoon five years ago when your approach to teaching a concept completely changed — and it cannot write what changed it, what you tried that failed, and why what you do now works so much better.
That specificity is what students remember. It is the difference between reading a good textbook and being in the room with someone who has done the thing. Textbooks are useful. Rooms are transformational. Your course should feel like the room.
The Five Things Only You Should Write
Your origin story — why you teach what you teach, and what personal experience made you the right person to teach it — should always be yours. Students buy from people they believe have been where they want to go. AI can write a version of your origin story, but it will be generic unless you have given it every detail. Write it yourself first, then let AI polish it.
Your frameworks and models need your fingerprints on them. If you have developed a three-step process, a diagnostic tool, or a decision framework through your own work with clients, write the explanation of it in your words. This is your intellectual property and your teaching signature. AI can help you refine and structure it, but the insight that produced it came from your experience.
Your direct coaching moments — where you are speaking to a student who is stuck, scared, or about to give up — need your authentic voice. AI coaching language is detectable and it lands flat. The moment a student most needs to hear “I know this feels impossible right now and here is what actually helps” should come from you.
Your genuine opinions and contrarian takes belong to you. If you think a commonly taught approach is wrong, overrated, or missing something important, that perspective is part of your value. AI will hedge. You should not.
Finally, your examples from your own client work — even generalised and anonymised — should be written by you. Real case studies carry a weight that hypothetical examples never do.
What This Means for Educators
Using AI to write the scaffolding — the intros, the summaries, the exercises, the frameworks explained in plain language — frees up your writing time for the five things above. That is a good trade. Students will not remember the lesson intro. They will remember the moment you told them what you actually believe.
The Simple Rule
If AI could have written it without knowing anything specific about you, let AI write it. If it requires something only you know, write it yourself.