I need to vent about this and also genuinely ask for advice because I’m kind of freaking out
Im a grad student. Just submitted a literature review that took me three weeks to write - like actually write, from scratch, using google scholar and my university library. No ChatGPT. No AI tools. Not even grammarly. Just me, a word processor, and too much coffee.
Got an email from my professor saying turnitin flagged it 48% AI generated and I need to “explain myself.”
I didn’t use AI. I have my google docs version history showing every single keystroke basically. But I’m not sure thats enough?? Has anyone succesfully challenged one of these flags??
from what I’ve read since this happened, turnitin’s AI detection has a known false positive problem especially with academic writing. apparently writing in a formal structured style with proper citations just… looks “too clean” to the algorithm. which is insane because isnt that literally what we’re trained to do in grad school???
this whole experience honestly radicalized me about AI detection. we need way better systems than “the algorithm says you cheated”
1 Like
Oh man, I feel this so hard. Same thing happened to my roommate last semester - her entire thesis proposal got flagged at 62%. She’s an ESL speaker who writes in a very formal, structured way because that’s how she learned English. Took her 3 weeks and a meeting with the dean to sort it out.
Google docs version history helped in her case. Also she showed all her research notes and early drafts. Basically had to prove her entire process which felt extremely invasive but it worked.
The false positive rate on Turnitin’s AI detection is embarassingly high for a tool that carries this much institutional weight. There was a study published a few months back (I’ll try to find it) showing something like 20% false positive rate on academic writing from non-native english speakers.
My advice: compile EVERYTHING. Version history, notes, browser history showing your research, timestamps. Go in with more evidence than they’re expecting. And push back hard — you have every right to.
Worth noting that Turnitin themselves say their AI detection shouldn’t be used as the sole basis for academic misconduct decisions. It’s literally in their documentation. Your professor should know this.
If your university is using it as a binary “guilty/not guilty” tool, that’s an institutional failure, not a you problem. I’d escalate above the professor if they dont respond well to the evidence.
i teach at a community college and i stopped using turnitin’s ai detection for exactly this reason. flagged 3 of my best students last year, all false positives, all incredibly stressful for them.
now i just use it for plagiarism checking (which it’s actually good at) and ignore the AI score completely
sorry you’re dealing with this. the system is broken.
update: met with my professor today. brought my google docs history, all my research notes, even showed my library checkout history lol. he was actually pretty understanding once he saw everything — said he’d had a few other false positives this semester and is reconsidering how the department uses the tool.
not marking it as resolved on my record at all which is a huge relief.
thank you all for the advice, seriously. @RushMoment if you find that study please share it, I want to send it to the department head