How to spot AI use and some tips to deal with this in class.
This Doesn’t Sound Like Them…” What Teachers Are Really Noticing About AI in Student Writing
You’re marking a stack of assignments. It’s been a long day. You’re halfway through your third coffee, and then you pause.
One essay catches your attention.
It’s good. Very good. Maybe too good.
The structure is clean. The vocabulary is precise. The transitions flow perfectly: Moreover, Furthermore, In conclusion… It reads like a model answer.
And that’s exactly the problem.
Because it doesn’t sound like your student.
That Moment of Doubt
Most teachers have had this moment now. Not because students are doing something dramatically wrong, but because something feels slightly off.
Not wrong. Just unfamiliar.
The mistake many teachers make at this point is jumping straight to tools. Running the text through detectors. Looking for a percentage score that confirms a suspicion.
But if you’ve been teaching for a while, you already know something important:
No tool knows your students the way you do.
AI detection doesn’t start with software. It starts with memory.
You remember how this student writes. The kind of sentences they build. The mistakes they repeat. The words they avoid. Their rhythm.
So when something shifts too quickly, you notice it, not because it’s perfect, but because it’s different.
When Writing Improves Too Fast
Progress is normal. Even rapid progress can happen.
But a complete transformation overnight? That’s when teachers pause.
It’s not just that the writing is better. It’s that it feels like it belongs to someone else. The tone changes. The sentence length changes. The confidence changes.
And the usual mistakes, the ones you’ve been correcting for weeks, disappear entirely.
That’s often the first quiet signal.
The “Perfect” Essay That Feels Impersonal
Then there’s the style.
AI writing tends to be very balanced. Almost too balanced. Every paragraph is neatly organised. Every idea is introduced, explained, and concluded with mechanical precision.
You start seeing the same phrases appear again and again:
Moreover In addition In conclusion
Sometimes you notice something else: two or three students submitting work that feels oddly similar. Not identical, but close enough that you start to wonder.
It’s not plagiarism in the traditional sense.
It’s more like they all asked the same invisible assistant for help.
When Writing Sounds Smart But Says Very Little
Another common moment: you read an essay that sounds intelligent, but when you finish, you realise you’ve learned almost nothing new.
The ideas are correct. Safe. Polished.
But also empty.
There’s no risk, no personal voice, no real engagement with the topic. It feels like a summary of what someone should say, rather than what this student actually thinks.
And then you check the references.
A source looks legitimate until you click it.
And nothing opens.
Or worse, it leads somewhere unrelated.
That’s when suspicion becomes a little stronger. Not certainty. But enough to look closer.
The Conversation That Reveals Everything
This is where everything becomes clearer, not through software, but through conversation.
You sit with the student and ask something simple:
“Can you explain what you mean in this paragraph?”
If the work is theirs, even if it’s imperfect, they can usually explain it. Maybe not perfectly. Maybe with hesitation. But the thinking is there.
If it isn’t, the response often sounds different:
Long pauses Vague explanations Repetition of the same sentence without real understanding
Sometimes they recognise the words, but not the meaning behind them.
That’s often the moment where the issue stops being about AI and starts being about learning.
The Invisible Process That’s Missing
Writing, in real life, is messy.
Students cross things out. They rewrite sentences. They hesitate, adjust, rethink.
So when a document appears fully formed, no drafts, no edits, no visible process, it raises questions.
Sometimes the answer is simple: the student copied and pasted something in one go.
And sometimes, they didn’t realise that this matters.
Because what we assess isn’t just the final product. It’s the thinking behind it.
What Most Teachers Realise Eventually
At this point, many teachers expect a clear answer. A moment where everything becomes obvious.
But that moment rarely comes.
Instead, what you get is a pattern:
A shift in style A lack of depth Difficulty explaining No visible process
Individually, none of these prove anything.
Together, they tell a story.
And Then Comes the Decision
What do you do next?
This is where teaching becomes more important than detecting.
Because in many cases, students are not trying to cheat in the way we traditionally understand it. They are experimenting. Testing boundaries. Using tools they don’t fully understand.
Some think AI is just another version of Google. Some think editing AI text counts as original writing. Some simply don’t realise they’ve crossed a line.
So instead of starting with accusation, the most effective teachers start with a question:
“Can you show me how you wrote this?”
And then they listen.
Where the Real Teaching Happens
This is the part that has changed the most.
Before AI, conversations about academic honesty were often reactive, triggered by clear cases of plagiarism.
Now, they’re becoming proactive.
Teachers are showing students:
How to use AI to generate ideas, not final answers How to rewrite content in their own voice How to question what AI produces When to acknowledge its use
Because the goal isn’t to eliminate AI.
It’s to make sure students are still doing the thinking.
A Shift in Perspective
At some point, many teachers reach the same conclusion:
This isn’t just about catching AI use.
It’s about recognising when learning is happening and when it isn’t.
A perfectly written essay without understanding is less valuable than an imperfect one that reflects real thinking.
And students, when guided properly, understand that.
Final Thought
You’ll probably keep having that moment:
“This doesn’t sound like them.”
And sometimes, you’ll be right.
But what matters more is what you do next.
Not the tool you use. Not the percentage score you get.
But the conversation you start and the way you guide your students from there.
