Direct answer
An AI detector in writing is software that analyzes a document's text and estimates how much of it was likely produced by a generative model such as ChatGPT, Claude, or Gemini. Turnitin's AI writing detection works inside the institutional similarity report, returning an overall percentage of qualifying text flagged as AI-generated together with sentence-level highlighting [1]. The result is a probabilistic signal for instructors to interpret, not a verdict on authorship [1].
How Does an AI Detector in Writing Decide Whether Text Was AI-Generated?
An AI detector does not "know" who typed a sentence. It measures statistical patterns in the text and compares them against patterns learned from large volumes of human and machine writing. Turnitin's model analyzes a submission and returns both an overall AI percentage and segmentation that marks which passages contributed to it [2]. Because the output is split this way, a document can be partly flagged rather than uniformly labelled.
Length matters as much as content. Only text that meets a minimum length threshold is included in the calculation, so short passages, headings, reference lists, and brief quotations are typically excluded from the percentage [2]. This is why two essays of very different lengths can produce surprising scores: the denominator is qualifying text, not the whole file.
It also matters what the percentage is not. It is the proportion of qualifying text the model considers likely AI-written — not the probability that a student cheated, and not a measure of intent or effort [2]. Detection is inherently probabilistic, and Turnitin advises treating the report as a signal that requires human judgment and context before any conclusion is drawn [1].
What Does the AI Percentage in a Turnitin AI Writing Report Actually Mean?
The headline number is a share, not a verdict. Turnitin reports the proportion of qualifying text that matches its model's AI-writing patterns, and it explicitly states that the score is not a measure of authorship or intent [3]. A high percentage means the model found a large share of qualifying text resembling machine output; it does not establish that a student used a chatbot improperly.
One detail confuses almost everyone: when the AI score falls below Turnitin's 20% confidence threshold, the report displays an asterisk (%) instead of an exact figure [3]. Those asterisks are low-confidence signals — the detector is not confident enough to publish a precise number. Seeing % is therefore materially different from seeing 60%, even though both appear in the same field of the report.
The practical reading is segment-first, number-second. Instructors are advised to review the highlighted passages and the surrounding context rather than acting on the headline figure alone [3]. A flagged paragraph may be a heavily templated methods section, a standard definition, or a passage the student drafted with AI assistance and then edited — all of which look different when read in context [2].
Can Students Check Their Own Writing With a Turnitin AI Detector Before Submitting?
Turnitin's institutional product is licensed to schools and universities, so student access is controlled by the institution rather than by individual purchase [4]. That means most students cannot simply log in and run a draft whenever they want; the ability to generate a report depends on whether their school has enabled it and on how the assignment is configured.
Where pre-submission checking is available, it is usually offered in a non-repository mode so a draft can be reviewed without being added to the institutional paper database [4]. Non-repository checking is the reason a rehearsal run does not contaminate later similarity results. Students who want a preview of both the AI and similarity reports often look for this option specifically, because it mirrors what the instructor will eventually see.
The timing pressure is real. Students typically seek a preview during the window between finishing a draft and the graded deadline, so turnaround and clarity matter as much as the raw score [4]. Understanding the report structure — overall percentage, asterisk threshold, and highlighted segments — is what makes a preview actionable rather than merely alarming [3].
Reading the report is only half the job; the other half is seeing your own draft through the same lens your instructor will use. That is exactly what turnitin0 was built for — students upload a.docx,.pdf, or.txt and receive the Turnitin AI detection report and the similarity report together, so the flagged segments and the score band are visible before the deadline rather than after it.
※ Turnitin0.com - Actual Turnitin AI Report Cover, Score, Flag And Similarity Summary
FAQ
Is an AI detector in writing the same as a plagiarism checker?
No. A plagiarism checker compares your text against existing sources and the student paper database, while an AI detector compares your text against patterns learned from machine-generated writing [1]. Turnitin surfaces both in one report, which is why students see an AI percentage and a similarity percentage side by side [2].
Why does my report show an asterisk instead of a number?
Turnitin displays % when the AI score falls below its 20% confidence threshold, meaning the detector is not confident enough to publish an exact percentage [3]. Treat % as a low-confidence signal rather than a confirmed result.
Can a human-written essay be flagged as AI-generated?
Yes. Because detection is probabilistic and based on statistical patterns, heavily templated or formulaic prose can resemble machine output [1]. That is why the report highlights specific segments and why instructors are advised to read them in context before drawing conclusions [3].
What text is excluded from the AI percentage?
Only text meeting a minimum length threshold is analyzed, so short passages, headings, and brief quotations generally fall outside the calculation [2]. This affects the denominator and can make scores hard to compare across documents of different lengths.
Can I see my own Turnitin AI report before I submit?
Access to Turnitin's institutional product is controlled by your school, not by individual purchase [4]. Where pre-submission checking is offered, it is usually non-repository, so a draft can be reviewed without entering the paper database [4].