Is it really true that Turnitin can detect AI-generated essays? It clearly states that it detects AI writing, but is that accurate? We all know Turnitin has long claimed it can detect papers purchased from essay writing services, though many students still avoid getting caught. What about AI detection, though? Has anyone tried submitting an essay generated with the help of ChatGPT? I'm swamped with essays and tempted to use AI, but I'm also scared af it'll get flagged since my school uses Turnitin.
Can Turnitin detect AI writing / ChatGPT?
- Thread starter NadiaT
- Start date
Well, yes and no. The truth is that there is no reliable way of detecting AI essays. Turnitin gives you an estimated % of the likelihood that a submitted paper was AI generated.
This percentage can sometimes be high even for those papers that were written from scratch. Here's a Reddit thread where a student says that Turnitin rated their essay as 62% AI-generated:
AI detectors often rely on patterns or characteristics that might be common in AI-generated text but can also appear in human-written papers. This can lead to A LOT of false positives. Plus, if someone mixes human and AI writing cleverly, it becomes even harder to detect. Turnitin's AI detector is definitely not foolproof.
This percentage can sometimes be high even for those papers that were written from scratch. Here's a Reddit thread where a student says that Turnitin rated their essay as 62% AI-generated:
AI detectors often rely on patterns or characteristics that might be common in AI-generated text but can also appear in human-written papers. This can lead to A LOT of false positives. Plus, if someone mixes human and AI writing cleverly, it becomes even harder to detect. Turnitin's AI detector is definitely not foolproof.
NerdAlert909
New member
So I recently turned in my final essay for a class, and my professor said Turnitin flagged it with a 23% AI-generated score. I told him I didn't use any AI at all. He kinda just shrugged and gave me an okay grade but told me to rewrite it.
I’ve been working on that rewrite and just checked it with ZeroGPT. Weirdly, this time it scored even higher at 34%. I've already emailed my professor with a screenshot of the new score.
Once he gets back to me, I’m thinking of sending him photos of all the books and notes I used. Other than that, I’m out of ideas. This AI detection thing seems pretty hit or miss, and it's putting students in a tough spot with potential false flags. If anyone has tips on how to prove I didn’t use AI, I’d really appreciate it!
I’ve been working on that rewrite and just checked it with ZeroGPT. Weirdly, this time it scored even higher at 34%. I've already emailed my professor with a screenshot of the new score.
Once he gets back to me, I’m thinking of sending him photos of all the books and notes I used. Other than that, I’m out of ideas. This AI detection thing seems pretty hit or miss, and it's putting students in a tough spot with potential false flags. If anyone has tips on how to prove I didn’t use AI, I’d really appreciate it!
Just take one of his papers and toss it into an AI checker. I bet it'll score as high as yours.
First off, that AI detector is really no good. It seems like it was rushed through development. Even big AI companies struggle to make these tools, so there’s no way Turnitin whipped up a reliable one so quickly.
Also, a 23% probability of content being AI-generated? That's super low.
Plus, you should try uploading your class readings to Turnitin. I bet it’s going to flag a bunch of that material as AI-generated too. I remember reading a similar story about a high school student whose essay was marked as AI-generated by their principal and teacher. Can't recall exactly where I saw it, though.
YourWriter
New member
I tried running three of my papers through ZeroGPT and scored between 20% to 30% on each one. These AI detectors are trash.
PuzzledPenguin
New member
I feel like students who use chatgpt to write their papers and then tweak them with other software to dodge AI detection are probably going to have an easy time. Meanwhile, students who actually write well but don’t use chatgpt might keep getting mistakenly flagged for AI use.