On not using turnitin

Originally published: October 23, 2025
Last updated: October 24, 2025

For many years, I have despised the use of TurnItIn¹. There are three main issues that I have with this:

1. Pitting teachers/professors against students
2. False Positives
3. Using my data to train the model TurnItIn is the software that my university uses, but there are many others out there.

   Adversarial relationship

   The TurnItIn approach pits teachers against students from the very beginning. It essentially says, “I can’t trust that you won’t plagiarize so you have to run your paper through this computer tool to make sure.”

Rather than teaching how to cite properly and put patterns in place to help students know beforehand, some teachers/professors have taken the approach that they must submit to TurnItIn before they even look at the paper.

False Positives

Plagiarism detectors detect things incorrectly. It’s not perfect, like most things aren’t. But false positives is a big problem. And it makes me not want to use it at all.

Vanderbilt University² reached the same conclusion after months of testing Turnitin’s AI detector, noting both its unreliability and its risk of false positives. Vanderbilt reported that if applied to its 75,000 submissions in a single year, nearly 750 students could have been falsely accused of AI misuse. Vanderbilt therefore disabled Turnitin’s AI detection feature “in pursuit of the best interests of our students and faculty”.

Using my data to train the model

According to the EULA, by submitting my dissertation I grant Turnitin “a non-exclusive, royalty-free, perpetual, worldwide, irrevocable license to use such papers… for improving the quality of the Services generally” This license survives termination of the agreement, meaning my dissertation would remain indefinitely in Turnitin’s private database and potentially be used to evaluate other students’ work.

As I don’t want my data used to further train the general LLM models for the companies I am using, I also don’t want TurnItIn to use my data to further their business pursuits.

Some would see using student data as a benefit that that student then won’t be plagiarized. As I have talked with professors, they have indicated they would want to make sure they aren’t accidentally plagiarizing something.

I have academic integrity, and tools like TurnItIn start with the assumption that I don’t. To me, using the tool is unethical for a professor to use.

An Alternative

I would like something where I can build my own checker, using my own files. I started doing that with Nexa’s Hyperlink, which is all local on your computer.

1. Turn it in is a tool that checks your work to make sure you’re not plagiarizing. Learn more at their web site ↩

2. Vanderbilt turned off the whole AI detector, as other schools have done, because of so many false positives. ↩

Notes mentioning this note

There are no notes linking to this note.