Direct answer
The Turnitin AI Flagger is a color-coded detection indicator within Turnitin's AI Writing Report that flags submitted documents based on the estimated percentage of AI-generated content. The flagging system uses three visual states—no flag for scores under 20% (shown as *%), a yellow indicator for scores between 20% and 80%, and a red indicator for scores above 80%—enabling instructors to rapidly assess whether a submission may have been written with AI assistance [1]. Alongside the overall flag, Turnitin provides sentence-level highlights that pinpoint specific passages the detection model considers likely AI-generated.
How Does the Turnitin AI Flagger Work?
The Turnitin AI Flagger operates as an integrated layer within the Feedback Studio report, analyzing submitted text against a detection model trained on academic writing and AI-generated content. When a document is processed, the model evaluates each sentence and the document as a whole, producing an overall AI score from 0 to 100 [2].
This score is then translated into three flag categories: no flag (0–20%), a yellow flag (20–80%), and a red flag (80–100%). The flag is displayed prominently on the report card alongside the similarity score, giving instructors a quick visual reference [2]. The flag is not a binary pass/fail—it reflects the model's confidence that portions of the text were produced by AI tools such as ChatGPT, Claude, or Gemini.
Beyond the overall flag, Turnitin's report highlights individual sentences or paragraphs that the model flags as likely AI-generated. This dual approach—document-level flag plus sentence-level highlights—allows instructors to examine flagged content in context rather than relying on a single number [2]. The flag indicator and highlights are visible to instructors when they open the AI Writing Report within Turnitin Feedback Studio.
What AI Score Triggers a Flag on Turnitin?
Turnitin uses three distinct flag thresholds based on the overall AI score calculated for a submission. These thresholds are built into the AI Writing Report and are not adjustable by individual instructors or institutions [3].
For scores between 0% and 20%, Turnitin displays no flag indicator. The system treats this range as unlikely to contain significant AI-generated content, and scores below 20% are shown as an asterisk (*%) rather than the exact number to avoid over-interpretation of low-confidence readings [3].
For scores between 20% and 80%, the report shows a yellow flag indicator. This indicates that the detection model found moderate evidence that portions of the document were AI-generated. Instructors are encouraged to review the sentence-level highlights alongside the overall score before drawing conclusions [3].
For scores between 80% and 100%, the report shows a red flag indicator. This signals that the model has high confidence that most of the document was AI-generated. A red flag typically warrants a closer conversation between instructor and student about the writing process [3].
Can Students Check Their AI Flag Score Before Submitting?
In standard Turnitin configurations, students generally cannot view the AI Writing Report—including the flag indicator and AI score—before their instructor opens it within the institutional LMS. Whether students can access the report after submission depends entirely on the settings configured by their school or university [4].
Turnitin does not provide a student-facing preview tool that allows learners to upload a draft and see their AI flag score before formal submission. This means that students who want to understand their AI detection risk must find alternative approaches [4].
One practical solution is to use a third-party Turnitin report service that mirrors what instructors see in their academic systems. These services generate the same AI Writing Report with the same flag indicators (no flag, yellow, red) and the same *% display for sub-20% scores, allowing students to preview their results before submitting to their institution [4].
If you want to see exactly where your draft stands with Turnitin's AI flagger—before your instructor does—Turnitin0 delivers the same AI Writing Report that professors see in their institutional systems, complete with the flag indicator and sentence-level highlights. No subscription needed, just upload your file and get your results in minutes.
※ Turnitin0.com - Actual Turnitin AI Report Cover, Score, Flag And Similarity Summary
FAQ
Does a yellow flag mean my paper was written by AI?
No. A yellow flag (20–80% AI score) indicates that Turnitin's model detected moderate evidence of AI-generated text, but it is not a definitive judgment. Instructors typically review the yellow flag alongside the sentence-level highlights and discuss the results with the student before making any determination [3].
Can the Turnitin AI flagger detect all AI writing tools?
Turnitin's model is trained to detect text from major LLMs including ChatGPT, Claude, and Gemini, but detection accuracy varies. The flag reflects the model's confidence level—not absolute certainty—and false positives are possible, especially with highly formulaic or structured academic writing [1].
Why does my score show *% instead of a number?
Turnitin displays any AI score below 20% as *% (asterisk) rather than a specific single-digit percentage. This is by design: the model's confidence below 20% is not statistically reliable enough to report an exact number, so the asterisk indicates minimal to no AI detection [3].
Can instructors see the AI flag before the similarity score?
Yes. In Turnitin Feedback Studio, both the AI indicator (flag) and the similarity score appear on the same report card. Instructors see the flag color immediately, alongside the percentage and sentence-level highlights, all within the same interface [2].
How long does Turnitin take to generate the AI flag?
The AI Writing Report is typically generated within minutes of submission, alongside the similarity report. In most institutional LMS integrations, the flag appears as soon as the report finishes processing, though instructor visibility settings may delay when students can see it [1].