Direct answer
An artificial intelligence detector is a tool that analyses a piece of writing and estimates how much of it was likely produced by a generative AI model such as ChatGPT, Claude, or Gemini. In higher education the detector most students actually meet is Turnitin's AI writing detection, which runs inside the same report that shows similarity matches [1]. It does not read the author's mind and it does not certify authorship; it returns a statistical signal about the text, expressed as a percentage of the document that appears AI-generated [1]. Understanding what that number does and does not mean is the difference between panicking over a flag and using it to improve a draft.
What Is an Artificial Intelligence Detector and How Does Turnitin's AI Writing Detection Work?
Turnitin's AI writing detection is not a separate product you have to request — it is generated automatically alongside the similarity report whenever a document is processed, and it highlights suspected AI-generated passages directly inside the document view [2]. The model was trained to recognise statistical patterns typical of large language models, so it is looking at how sentences are constructed rather than at whether the ideas are correct or the citations are real [1].
There is an important mechanical limit: Turnitin only produces an AI detection result once a minimum amount of text has been analysed, roughly 300 words, because short samples do not give the classifier enough signal to work with [2]. Quotations, reference lists, and bibliographies are handled differently from body prose, which is why a heavily quoted literature review can look cleaner than the argument around it [2].
The practical consequence is that an artificial intelligence detector is a screening instrument, not a verdict. Turnitin itself frames the indicator as one piece of evidence that should sit alongside context, drafting history, and a conversation with the student — not as standalone proof of misconduct [1]. Treating a single percentage as conclusive is a misreading of what the tool was built to do.
What Do Turnitin's AI Score Bands and the *% Result Actually Mean?
The headline number in the report is the share of the analysed document that the detector flags as AI-generated, and Turnitin presents it as a percentage so that a 40% result means roughly two-fifths of the analysed text matched AI-generation patterns [3]. That percentage is about the text, not about intent: a student who wrote every word themselves can still trigger a flag if their prose happens to resemble model output.
The single most misunderstood detail is the asterisk. When detection confidence falls below Turnitin's threshold, the report displays *% instead of an exact figure, and that threshold sits at 20% — meaning anything in that zone is a low-confidence signal rather than a firm AI percentage [3]. Students frequently read *% as "20% AI" when it actually means the classifier could not commit to a precise value.
Score bands are best read as ranges rather than pass/fail lines. A low band suggests little AI-like patterning, a mid band suggests mixed evidence worth reviewing, and a high band indicates that a large share of analysed text matched AI patterns and deserves close attention before submission [3]. In every band, the useful next step is the same: read the highlighted passages and decide whether they genuinely represent how you write.
Can Students Preview a Real Turnitin AI and Similarity Report Before Submitting?
Turnitin positions student-facing report access as formative — the point is to review flagged passages and revise weak or derivative writing before it is graded, not to reverse-engineer a score [4]. In practice, though, access is uneven: many students only see a report after an instructor enables it in the LMS, and some never see the AI indicator at all, so a pre-submission preview is not guaranteed through institutional channels [4].
That gap is exactly why independent pre-checks exist. A non-repository check lets you see the same two artefacts your instructor sees — an AI detection report and a similarity report with match highlights — while your file stays out of Turnitin's student paper database. Turnitin's own guidance encourages students to use report feedback to improve citation practice and originality, and to bring questions to their instructor rather than treating the number as a final judgement [4].
If you want that preview to be meaningful, run it on the version you actually intend to submit, and read the flagged passages rather than only the headline percentage. A report you understand is far more useful than a number you merely fear.
Reading the bands is one thing; seeing your own draft scored against the same report your professor opens is another. turnitin0 exists for that moment — a pre-submission check that returns the AI detection report and the similarity report together, so you know exactly what you are walking into before the deadline does the deciding.
※ Turnitin0.com - Actual Turnitin AI Report Cover, Score, Flag And Similarity Summary
FAQ
Is an artificial intelligence detector the same as a plagiarism checker?
No. A plagiarism checker compares your text against existing sources to find similarity matches, while an AI detector estimates whether the prose itself looks machine-generated [1]. Turnitin produces both in the same report, which is why students often see an AI percentage sitting next to a similarity percentage.
Why does my Turnitin AI report show *% instead of a number?
An asterisk appears when the detector's confidence falls below its threshold, which Turnitin sets at 20% [3]. It signals a low-confidence result rather than a precise AI percentage, so it should not be read as a hidden exact score.
Can a human-written essay still be flagged by an AI detector?
Yes. The classifier responds to statistical patterns in sentence construction, so formal, formulaic, or heavily templated academic prose can resemble model output even when a person wrote every word [1]. That is precisely why Turnitin describes the indicator as one signal among several rather than proof [1].
How much text does Turnitin need before it will detect AI?
Turnitin requires a minimum amount of analysed text, roughly 300 words, before it returns an AI detection result [2]. Very short excerpts therefore may produce no AI indicator at all.
Where can I see the same report my instructor sees?
Institutional access depends on whether your instructor has enabled the report in the LMS, and many students only see it after submission [4]. A non-repository pre-submission check lets you review the AI and similarity reports on your own draft before the final upload.