With more Nigerian universities running submissions through Turnitin's AI detection feature alongside its usual plagiarism check, this question comes up constantly, and honestly, the answer is more than a simple yes or no. Let's go through what Turnitin's AI detection actually does in 2026 and what that means for how you should be using AI tools on your project.
I need to tell you that beyond the traditional similarity report most students are already familiar with, Turnitin also runs a separate AI writing detection score on submitted documents, estimating how much of the text appears to have been generated by tools like ChatGPT. Turnitin rolled out a major update to this system in January 2026, specifically aimed at catching text that had been run through paraphrasing or humanizing tools designed to disguise AI generated writing.
Turnitin reports a high accuracy rate, in the high nineties as a percentage, with a low false positive rate, for documents where a substantial portion of the content is AI generated. But independent research paints a more mixed picture, especially once AI generated text has been manually edited or paraphrased, where detection rates drop considerably. Some studies have also found that the tool is more likely to flag writing from non native English speakers or highly formal, structured academic writing as suspicious, even when it was written entirely by a human.
The safest way to think about this is that Turnitin's AI score is a probability estimate, not solid proof, and universities are increasingly being advised to treat it that way rather than as automatic evidence of misconduct. Still, that does not mean a high AI score is harmless, most departments treat a flagged submission as something that triggers further questions, sometimes an uncomfortable conversation with your supervisor, even if it does not automatically mean failure. Getting flagged at all is a situation worth avoiding entirely if you can.
If you use AI tools at any stage of your project, for brainstorming, outlining, or getting unstuck on a difficult section, the risk is genuinely low. The risk rises sharply when large sections of your actual submitted writing are AI generated with little to no editing in your own voice, since that is exactly the pattern that current detection systems are built to catch.
Turnitin updates its detection model fairly regularly, and universities themselves are still adjusting their own policies around how strictly to treat AI scores, with some institutions elsewhere in the world choosing to disable AI detection entirely over reliability concerns. What is true about the technology and the policies around it today may shift again within the next year, so it is worth checking your own department's current stance rather than assuming it matches what you have heard from other students or other schools.
Assuming a high AI detection score is automatic proof of misconduct with no room for explanation
Assuming AI detection tools are perfectly accurate and cannot produce false positives
Submitting large sections of unedited AI generated writing and assuming it will pass unnoticed
Not checking their own department's specific current policy on AI tool use
Avoiding all use of AI tools out of fear, even for legitimate research support like brainstorming or outlining
You need to understand how this technology actually works, rather than relying on rumors from other students. This will put you in a better position to use AI tools sensibly (to brainstorm) without unnecessary risk to your project. If you want your project properly written, reviewed, and quality checked from the start, my team at ProjectPal have built every project around genuine originality. Reach out to us on WhatsApp to get started.
Send your topic and department on WhatsApp and get support built around your actual project.