Direct answer
An AI text detector is software that estimates the likelihood a passage was generated by a large language model rather than written by a human, and the version most university students actually meet is Turnitin's AI writing indicator, which runs inside the similarity report [1]. It returns a percentage of the analysed text it considers AI-generated, not a verdict on authorship or quality [1]. Because that number is probabilistic, the score matters less than the band it falls into and the passages it flags — which is why reading the report correctly is the real skill [1].
What does an AI text detector actually measure, and how accurate is it?
An AI text detector does not test for plagiarism, and it does not judge whether your argument is good. Turnitin's model analyses statistical patterns across segments of your prose and estimates how much of the writing looks machine-generated, then reports that as a share of the document [2]. In practice this means a detector is measuring predictability and structure — the smooth, evenly weighted sentences that language models tend to produce — rather than searching for a copied source [2].
Accuracy is where expectations usually break. Turnitin publishes its detector as a signal to support academic judgement, not a replacement for it, and it evaluates text in segments so that a single flagged paragraph can pull a whole document's percentage upward [2]. A high score on a document you drafted yourself is therefore possible, and it is the reason false-positive discussion sits at the centre of every serious detector guide [2]. The honest framing is that a detector gives you a probability estimate with a confidence threshold attached, not a binary answer [2].
That threshold is the detail most students miss. Turnitin will not print a precise figure when its AI indicator is below its 20% confidence level, so the tool itself is telling you when it is not confident enough to commit to a number [3]. Treat any detector output as one input among several — your drafting history, your notes, and your supervisor's knowledge of your writing all carry weight [1].
What do the AI score bands and the asterisk (*%) result mean on a Turnitin AI report?
When you open a Turnitin AI report, the headline figure is the percentage of qualifying text flagged as AI-written, and the report breaks that down by segment so you can see exactly which sentences triggered the signal [3]. The bands are colour-coded for a reason: they let an instructor scan a cohort quickly and decide where to look more closely, rather than treating every number as equally serious [3].
The asterisk is the part that causes the most confusion. Turnitin displays an asterisk () in place of a numeric percentage when the AI writing indicator falls below its 20% threshold, which means the detector found some signal but not enough to state a confident figure [3]. A score of % is therefore a low-confidence result, not a hidden high score and not a clean bill of health — it is the tool declining to overstate its case [3].
If your report shows *%, the practical reading is that only a small share of your text raised any AI signal at all, and the flagged segments are worth reviewing on their own merits [3]. If you see a higher numeric band, the report's segment view tells you whether the flag is concentrated in one paragraph you can rewrite or spread across the document [3]. Either way, the number is a starting point for revision, not a charge sheet [1].
How can I preview the exact Turnitin AI report my professor will see before I submit?
The most useful thing you can do with a detector is run it before the deadline, while revision is still cheap. Turnitin's own guidance to institutions is that students should understand what the report shows and use flagged passages as a prompt to revise rather than as an accusation to defend against [4]. That only works if you see the report early enough to act on it.
There is a structural problem here, though. Turnitin's AI and similarity checking normally runs inside an institution's learning management system, which means students usually cannot trigger a check on a draft — they see the report for the first time when the submission is already final [4]. Turnitin's recommended alternative is transparency: instructors discussing AI writing results openly with students so the report becomes a learning tool instead of a surprise [4].
That gap is exactly what a pre-submission preview closes. Turnitin0.com runs your draft through the same Turnitin AI detection and similarity checking and returns both reports to you, so you can read your own score bands, inspect the flagged segments, and revise before anything is submitted [4]. With 100,000+ reports delivered and a 4.9/5.0 student satisfaction rating, it is built for the moment right before you press submit — and it is a non-repository check, so your file is not added to Turnitin's student paper database.
Reading about score bands is useful, but the only way to know what your own document scores is to see your own report. Turnitin0.com gives you the actual Turnitin AI and similarity PDFs — the same output your professor opens — in under 15 minutes in 98% of cases, so you still have time to rewrite the flagged paragraphs before the deadline rather than after it.
※ Turnitin0.com - Actual Turnitin AI Report Cover, Score, Flag And Similarity Summary
FAQ
Is an AI text detector the same as a plagiarism checker?
No. A plagiarism checker matches your text against source documents, while an AI text detector estimates whether the prose itself looks machine-generated [2]. Turnitin runs both inside the same report, which is why the two scores can move independently [1].
Why does my Turnitin AI report show an asterisk instead of a number?
The asterisk appears when the AI writing indicator is below Turnitin's 20% confidence threshold, meaning the result is low-confidence rather than a precise percentage [3]. It is not a hidden score — it is the detector declining to commit to a figure [3].
Can I check my own draft before submitting it to my university?
Institutional Turnitin access normally only covers final submissions, so students rarely get a pre-submission look [4]. A non-repository pre-check lets you read the same AI and similarity reports your professor will open, while you can still revise [4].
Does a high AI score mean I will be accused of misconduct?
Not on its own. Turnitin presents its AI indicator as a signal to support academic judgement, and institutions are advised to discuss results with students rather than treat them as automatic misconduct [4]. A high band is a prompt to review the flagged segments [1].
What file types and limits apply to a pre-submission check?
English documents in.docx,.pdf, or.txt work, with a word count above 300 and below 30,000, and files under 20 MB. Each check returns both the AI detection report and the similarity report as downloadable PDFs.