Is AI Good for Exam Prep or Just Assignments?
Hot take: AI is better for exam prep than it ever was for assignments.
Not just “for some people.”
For basically everyone.
Because exam prep forces the thing uni is meant to reward: learning.
Most of the AI panic isn’t about learning.
Often it’s all about submission.
The people who set the rules know we’ll use AI.
So heaps of effort goes into containing it—stopping cheating, and protecting the old methods of teaching.
That pressure lands hardest on assignments.
Exam prep doesn’t carry the same baggage.
So let’s split it up:
How AI Is Good For Assignments (When You Don’t Outsource Thinking)
You know the assignment moment:
Blank doc open, task sheet open, and you’re bouncing between the rubric and your notes like the phrase “critically evaluate” is going to suddenly explain itself.
AI can be great here… as long as it’s doing what a good tutor would do: clarify, structure, check—not write your brain for you.
Where it helps most
Decode the rubric: Turn the criteria into a simple checklist.
Build the structure: Outline first so you’re not patching chaos at 1am.
Unstick confusion fast: Re-explain concepts the second you stall.
Quality control + polish: Flag weak reasoning and tighten clarity.
The trap
If AI generates the reasoning or paragraphs, you’re “finishing” the assignment without building the skill it was meant to train.
And yeah… when the due date is in hours, it gets tempting.
How AI Is Good For Exam Prep (This Is Where It Shines)
Exam prep is a different game.
Because in the exam, there’s no ChatGPT tab, no “can you hint me through this,” and no “rewrite my answer so it sounds smarter.”
It’s just what you can actually pull from memory.
That’s exactly what AI is good at exposing—and fixing.
Where it helps most
Active recall drills: Turn notes into questions. You answer from memory.
Marks your attempts: Check your method against the rubric and show what you missed.
Gap-finding fast: Spot what you can’t retrieve cleanly.
Repeat loops: New similar questions until it sticks.
Here’s the main difference:
Assignments are easier to “complete” without learning. Exams aren’t.
So in practice:
Assignments: more restrictions, disclosure requirements, higher risk.
Exam prep: usually more acceptable, because it’s tutoring, not submission.
And I’m not here to lecture you about using it “the right way.”
You can make your own call there.
I’m just saying this:
Exams reward retrieval under pressure:
AI can’t sit the exam for you—so used for prep, it naturally forces:
Recall → feedback → correction → repeat.
The Verdict
So what’s the verdict this time?
I could give the standard answer:
“AI is good for both… it depends how you use it”… bla bla bla.
But too many people are thinking about AI wrong.
Like I said earlier: the system is mostly trying to contain AI where it can.
So we can all stick to the old methods of teaching.
But NO!!
We’re here to flip the script, not follow it.
If the goal is truly learning (as it should be), then when you compare AI for assignments vs exam prep… it’s pretty clear what comes out on top.
I’ve said it before and I’ll say it again:
The biggest gain with AI is the speed and depth of learning it can unlock—way beyond what most of us get access to otherwise.
Stay autonomous out there.
– The Prompted Learner Newsletter