Direct answer
A combined plagiarism and AI check runs two independent analyses on the same file: a similarity report that matches your text against published sources and a student paper repository, and an AI writing report that estimates how much of the document reads as machine-generated [1]. Turnitin presents these as two separate indicators — an overall similarity percentage with a colour-coded match overview, and an AI writing percentage with highlighted suspect passages [2]. Because institutional accounts usually restrict self-submission, most students run this preview through an independent checker before the real deadline [3].
What does a combined plagiarism and AI check actually return on my paper?
A plagiarism and AI check returns two distinct artefacts, not one blended number. The similarity side gives you an overall percentage plus a match overview that lists every source, colour-codes the overlapping text, and lets you click through to see exactly which sentences triggered a match [2]. The AI side gives you an overall AI writing percentage and an overlay that highlights the specific segments the model flagged as likely machine-generated [1]. Neither figure is a verdict on its own — they are signals designed to be read together.
The similarity percentage is the most misread number in academic integrity. It counts every matched string, including correctly cited quotations and bibliography entries, so a paper with heavy but legitimate referencing can show a high figure without any misconduct [2]. Most report views let you exclude quotes and references to see the underlying overlap, and that adjusted view is the one that actually tells you whether your paraphrasing is too close to a source [2]. A match against a previously submitted student paper is treated the same way as a match against a journal article, which is why the repository setting matters so much [2].
The AI writing indicator works on a different logic entirely. It reports an overall percentage of the document that the model considers likely AI-generated, then highlights those spans directly in the text [1]. When the detected signal is weak, Turnitin suppresses the precise figure and displays *% instead, because a low-confidence reading is not the same as a clean result [1]. Reading the AI indicator in isolation, without the similarity report next to it, is how students end up over-correcting a document that was never actually flagged [1].
Can I check my own paper for plagiarism and AI before submitting it to my university?
For most students, no — you cannot self-submit to your institution's Turnitin account. Submission access is controlled by the instructor's assignment settings, so if your tutor has not opened a draft or practice folder, there is no route in through the institutional system [3]. Even where a draft folder exists, the assignment configuration may still prevent your file from being added to the student paper database, which changes what the report can and cannot show you [3].
This restriction is precisely why independent pre-submission checkers exist. A third-party service runs the same two analyses on your draft and hands back downloadable reports covering both similarity and AI writing, so you see the shape of the result before the version that counts [3]. The practical value is timing: you get the numbers while there is still room to revise, rather than discovering a flagged paragraph after the deadline has passed [3].
One detail worth checking before you upload anywhere is whether the checker adds your file to a shared student repository. Institutional submissions sometimes do, which can create a self-match when you later submit the same text through your university account [3]. A non-repository check avoids that trap entirely, because your draft is analysed and then discarded rather than indexed [3].
How do I fix a high AI or similarity score before I submit?
Fix the underlying writing, not the number. Reducing a flag is a revision task: add specific evidence and concrete detail, vary your sentence rhythm, and strip out the generic connective phrasing that makes prose read as machine-generated [4]. Students who try to game the indicator with synonym swaps or inserted typos usually make the writing worse and the score no better, because the detector is responding to structural patterns rather than individual words [4].
For similarity, work through the match overview source by source. Where a match is a correctly cited quotation, leave it alone — that is expected academic practice [2]. Where a match is your own paraphrase sitting too close to the original, rewrite the sentence from your notes rather than from the source text, and check that the citation still points to the right place [2]. Excluding quotes and references in the report view first tells you how much real overlap you are actually dealing with [2].
For AI flags, the revision has to preserve what makes the paper yours. Keep your citations, headings, and argument intact while rewriting the flagged passages in your own voice, and do not let a rewrite strip out the evidence you gathered [4]. Because the AI indicator highlights specific spans, you can target those passages instead of rewriting the whole document [1]. The only way to confirm the flag actually dropped is to re-check the revised draft — a rewrite you never re-measure is just a guess [4].
Reading the report is the easy part; the harder question is whether the version you are about to submit is the version your tutor will see. turnitin0 was built for exactly that gap — a pre-submission check that returns the Turnitin AI writing report and the similarity report together, so you are revising against real numbers instead of a hunch.
※ Turnitin0.com - Actual Turnitin AI Report Cover, Score, Flag And Similarity Summary
FAQ
Does a plagiarism check also detect AI writing?
Not by default — they are two separate analyses. A similarity report matches your text against existing sources, while the AI writing indicator estimates how much of the document reads as machine-generated [1]. You need both reports to get a full picture of your draft.
Why does my AI score show *% instead of a number?
Turnitin displays *% when the detected AI signal falls below its confidence threshold [1]. It means the model found weak evidence rather than none, so treat it as a low-confidence reading and read it alongside the highlighted passages in the report [1].
Can my university see that I used a pre-submission checker?
An independent check runs on its own and is not reported to your institution. Choosing a non-repository check also means your draft is not added to a shared student paper database, so it will not create a self-match when you submit the final version [3].
What similarity percentage is too high?
There is no universal cutoff. A high figure driven by correctly cited quotations and references is normal, which is why report views let you exclude those elements before judging real overlap [2]. Look at the match overview source by source rather than reacting to the headline number [2].
How do I know my revisions actually worked?
Re-check the revised draft. The AI indicator highlights the specific spans it flagged, so you can confirm whether those passages still trigger a match after rewriting [1]. Re-measuring is the only way to know whether a revision genuinely lowered the score [4].