Overview:
This is a weekly short story series that focuses on teachers’ crazy stories—but with a little fiction in them to protect the innocent (and guilty). This week, it went down between a teacher, AI and her students, all over a Chemistry exam.
I’ve taught chemistry at Eastwood High for thirteen years. Long enough to tell when something doesn’t add up—even when it looks perfect on paper.
It started last Tuesday, after I finished grading the midterms. Normally, I’d be overjoyed when students scored high, but this time, something was off—way off.
Three of my juniors: Dustin, Eliana (who everyone calls El), and Mike, all scored perfectly on one of the most challenging exams I’ve ever given. Not just that—they wrote the exact same phrasing on multiple open-ended questions.
Not similar. Identical.
Every term, every comma. Even the extra credit question—”How does quantum tunneling influence chemical reaction rates? “which I tossed in last-minute and never taught in class, was answered with textbook clarity by all three.
And I mean textbook in the most literal sense. Their responses didn’t sound like teenagers. They sounded like polished dissertations.
Or worse… AI.
The Principal Doesn’t Want Trouble
I ran their answers through AI detection tools. They lit up like a Bunsen burner. Some results were flagged as “possibly generated,” but nothing definitive. So, I did what I’m supposed to do—I brought it to Principal Harrow.
He looked at me over his glasses and said, “You really think they cheated?”
I handed him the tests. “Something’s not right.”
He flipped through the papers. “But these are good students. El is top of her class. Dustin practically lives in the science wing. Mike gives campus tours. Unless we have hard proof, this could blow back.”
I left his office frustrated but not defeated.
Parents in Denial
Next, I called the parents.
Mrs. Henderson—Eliana’s mom—was stunned. “Eliana? No way. She’s always studying. She tutors half the junior class. You must be mistaken.”
Mr. Wheeler, Mike’s dad, was borderline offended. “Mike is brilliant. If anything, people cheat off him.”
And Dustin’s mom? “He loves chemistry. Our entire kitchen looks like a science lab.”
I got nowhere.
Lunchroom Interrogations
So, I went rogue.
I pulled each student in during lunch.
Dustin was first. Nervous energy. Fidgety. Talked a mile a minute about flashcards, YouTube channels, study groups. He said he stayed up three nights straight prepping. He looked like he hadn’t slept.
Eliana came next. Calm. Composed. Too composed. When I asked about the quantum tunneling question, she blinked—just once—then said, “I read about it… on Reddit?” It didn’t feel right.
Mike strolled in like he owned the place. Confident. Casual. Said he used Notion to organize his study materials and even showed me notes full of diagrams, footnotes, and citations. They were almost too good. No 17-year-old I know color-codes citations.
The Trap
I had one last trick.
I scheduled an unannounced after-school lab practical. Same material. Closed notes. Monitored room. Let’s see who knew the material without a screen feeding them answers.
Only Dustin and Eliana showed.
Dustin was visibly shaking, but he pulled through. Not flawlessly—but authentically. Real effort. Real confusion. Real learning.
Eliana… cracked. Her explanations were vague, hesitant. She struggled to articulate the difference between activation energy and reaction rate. I watched her confidence drain with every question.
Mike emailed fifteen minutes before the lab. Said he had the flu. Attached a doctor’s note.
In. Comic. Sans.
Signed by someone named Dr. Marcus Gatson.
I nearly laughed out loud.
The Whodunit
But here’s the problem. I still can’t prove anything. No digital footprint. No confession. And technically, no rule that says, “don’t feed test questions into ChatGPT.”
So now, I turn to you.
Three students. One AI-assisted cheat.
Was it:
- Dustin, the jittery science nerd with mysteriously perfect answers?
- El, the brilliant perfectionist who suddenly froze under real pressure?
- Or Mike, the smooth talker who got “sick” right before he had to prove himself?
You’ve seen the evidence. Now I want your hypothesis.
Who used AI to cheat—and how did they pull it off?
The lab is open. Your turn to investigate.
