Direct answer
Yes — Turnitin runs an AI writing check, and it is delivered inside the same report that carries your similarity score, flagging the share of your prose the model judges likely AI-generated [1]. That figure is a review signal, not a verdict: Turnitin positions the indicator as something an instructor reads alongside the highlighted passages, not as standalone proof of misconduct [1]. If you want to see that number on your own draft before the graded submission leaves your hands, you can preview it — and the rest of this guide explains what the score means, why it sometimes appears as *%, and how to check your work first.
Does Turnitin check for AI, and what does the AI score it gives you actually mean?
Turnitin's AI writing detection is not a separate product bolted onto the side of its plagiarism tooling. It surfaces within the similarity report workflow, and when your institution has the feature switched on, your instructor sees an AI indicator next to the familiar similarity percentage [2]. The two numbers answer different questions: similarity asks how much of your text overlaps existing sources, while the AI indicator asks how much of your qualifying prose the model considers likely machine-generated [2].
The headline AI figure is a proportion, not a probability of cheating. It represents the share of your analyzed text that was flagged as AI-generated, so a document written partly by hand and partly with a model can land at a partial percentage rather than all-or-nothing [2]. Turnitin also excludes material that should not count against you — short fragments, quoted passages, references, and bibliographies are stripped out of the calculation, which is why a heavily cited literature review is not automatically penalised for its citations [2].
Reading the number correctly matters more than reacting to it. A low band means very little of your analyzed prose was flagged; a high band means a large share of it was [2]. Turnitin's own guidance tells instructors to open the highlighted segments and judge them in context, because the percentage describes a pattern in the text, not the intent of the person who typed it [2]. In practice, that means the most useful thing you can do with an AI score is look at which sentences triggered it and decide whether they genuinely read as machine-written.
Why does Turnitin sometimes show an asterisk instead of a percentage in the AI report?
When the AI report shows *% rather than a number, you are looking at Turnitin's confidence threshold at work. The asterisk replaces a precise percentage whenever AI detection falls below Turnitin's 20% confidence threshold, which is the point at which the model stops committing to an exact figure [3]. In other words, the detector found something, but not enough flagged text to state a reliable number.
The asterisk is frequently misread in both directions. It is not a zero, and it is not a clean bill of health — it is a low-confidence signal, meaning the detected AI-flagged text is a small share of the document or the signal itself is weak [3]. Turnitin only publishes an exact percentage once a result reaches or exceeds that threshold, so anything under it collapses into the asterisk display [3]. Students who see *% and assume they are in the clear are reading the symbol as a grade when it is actually a confidence statement.
The practical takeaway is that *% tells you the detector is not confident enough to accuse your draft of being substantially AI-written. That is a meaningfully better position than a high exact percentage, but it still means some text caught the model's attention [3]. If your institution treats any AI flag as a conversation starter, knowing where that faint signal sits in your draft is more useful than knowing the symbol appeared at all.
Can students run a Turnitin AI check on their own draft before submitting it?
Sometimes — and it depends entirely on your instructor. Some instructors enable a draft or self-check folder inside the Turnitin assignment, which lets you submit early and read your own report before the graded attempt; many do not enable it at all [4]. Even where self-check exists, the default configuration can still add your file to the repository, so the draft you used to test may become part of the institutional paper database unless the instructor has deliberately set the assignment to non-repository [4].
That gap is why so many students search for an independent pre-submission check. When a draft folder is unavailable, an external check that returns the same two report types — an AI writing report and a similarity report — lets you read your own results and revise before the version that counts is uploaded [4]. The value is not in getting a different verdict from your institution; it is in seeing the same class of signal early enough to act on it.
A pre-submission check is most useful when you treat it as a revision tool rather than a scoreboard. If the AI indicator highlights specific paragraphs, you can rewrite those passages in your own voice, add your own analysis, and re-check; if the similarity report shows matches you had not noticed, you can quote and cite properly [4]. Either way, the point of checking first is that you still control the draft that gets submitted.
If you have read this far, you already know the uncomfortable part: Turnitin's AI indicator is real, it is attached to your submission, and *% is not the reassurance it looks like. What you cannot do inside your institution's assignment is choose when to see it — that decision belongs to your instructor. turnitin0 exists to hand that decision back to you: upload your .docx, .pdf, or .txt draft, and within minutes you get the same two reports your professor would see — the AI writing report and the similarity report — without your file entering Turnitin's student paper database. Over 20,000 students across the US, UK, Canada, Australia, New Zealand, and Ireland have used it to look before they leap, and 100,000+ reports have been delivered. Read your own AI score while there is still time to change it.
※ Turnitin0.com - Actual Turnitin AI Report Cover, Score, Flag And Similarity Summary
FAQ
Does Turnitin check every document for AI?
The AI writing indicator appears when your institution has the feature enabled and your instructor's assignment surfaces it, and it is delivered alongside the similarity report rather than as a separate submission [1]. If your instructor has not enabled it, you may see only the similarity score.
Is a *% AI score good?
It means AI detection fell below Turnitin's 20% confidence threshold, so the model will not publish an exact percentage [3]. It is a low-confidence signal rather than a zero, so it is better than a high exact score but not proof that no text was flagged.
What does a high Turnitin AI percentage actually prove?
Turnitin treats the percentage as a signal for human review, not as a determination of misconduct, and instructs instructors to inspect the highlighted passages in context [2]. The number describes how much of your analyzed prose was flagged, not why it was written that way.
Can I see my own AI report before the graded submission?
Only if your instructor has enabled a draft or self-check folder in the assignment [4]. Otherwise students typically use an independent pre-submission check that returns an AI report and a similarity report without adding the file to Turnitin's student paper database [4].
Which parts of my paper are excluded from the AI calculation?
Short fragments, quoted material, references, and bibliographies are excluded, because only prose of sufficient length qualifies for analysis [2]. That is why a citation-heavy section does not automatically inflate your AI percentage.