Direct answer
Direct Answer - Yes, Turnitin does flag AI-generated content. Its AI writing detection feature analyzes submitted academic text and produces an indicator showing the percentage of the document that was likely written by an AI tool such as ChatGPT, Claude, or Gemini. This flag appears as a score in the AI writing report, which instructors can view alongside the standard Similarity Report. Understanding how this flagging works — what scores trigger it, how it's displayed, and whether students can preview it before submission — is essential knowledge for anyone navigating modern academic integrity requirements [1].
How Does Turnitin's AI Detection Flag Content?
Turnitin's AI writing detection does not simply label a paper as "AI" or "not AI." Instead, it provides a nuanced indicator that instructors can interpret within the context of each submission. When a paper is processed, Turnitin analyzes sentence-level patterns and produces an overall percentage of AI-generated content [2].
The AI writing report presents this information in two complementary ways. First, it shows an overall percentage for the entire document — this is the primary "flag" that instructors see at a glance. Second, the report includes sentence-level highlights, marking specific passages that the model identifies as likely AI-generated, so instructors can see exactly which portions triggered the detection [2].
It is important to note that Turnitin's detection model is specifically trained on academic writing samples, which helps minimize false positives. The system is designed to be conservative — it will only flag text when it has a high degree of confidence that the content was produced by an AI tool rather than by a human writer. Instructors are trained to review the flagged portions holistically rather than treating any single percentage as definitive proof of misconduct [1][2].
What AI Score on Turnitin Gets Flagged as AI-Generated?
Turnitin does not use a simple binary "flagged" or "not flagged" system. Instead, the AI writing report communicates results through a percentage scale with specific display rules. Any score at or above 20% is shown as a precise numeric percentage — for example, 45% or 78% — giving instructors a clear measure of the AI content detected in the submission [3].
Scores below 20% are a different matter. In these cases, Turnitin displays the result as *% (an asterisk) rather than as a single-digit percentage like 3% or 12%. This design choice reflects Turnitin's position that low-confidence detections are not reliable enough to report as exact figures. The only explicit low numeric outcome a student typically sees is 0%, which indicates no AI content was detected at all [3].
For instructors, then, any score that appears in the report — whether as *% or as a specific percentage — serves as a flag that warrants a closer look. A *% result does not necessarily mean the paper is problem-free; it simply means the detected AI content fell below the 20% threshold. This is a critical distinction for students who may assume that a non-zero result is irrelevant [3].
Can Students Preview Their Turnitin AI Score Before Submitting?
Turnitin's AI writing detection report is typically available only to instructors after a paper has been submitted through an institutional learning management system. Students do not have direct access to the AI writing report through their university's Turnitin integration in most cases, which creates a challenge for those who want to check their work before the final submission [4].
However, third-party services have stepped in to fill this gap. Turnitin0.com, for example, allows students to upload their drafts and receive genuine Turnitin AI detection and similarity reports before they ever submit to their institution. This means students can see exactly what their AI score is — whether it shows *%, a specific percentage, or 0% — and make informed decisions about revising their work before the official hand-in [4].
The ability to preview an AI score before submission is not just about avoiding academic penalties. It also supports honest self-assessment and helps students understand how their use of AI tools is perceived by detection systems. Many educators recommend that students engage with these tools proactively as part of developing responsible AI literacy, rather than discovering the results only after a grade is at stake [4].
Want to know exactly where your draft stands before your instructor sees it? With Turnitin0, you can upload your paper and receive a real Turnitin AI report — including the AI score, flagged sentence highlights, and the full similarity report — all before you submit to your university. No subscriptions, no data archiving, and results delivered in about 10 minutes.
※ Turnitin0.com - Actual Turnitin AI Report Cover, Score, Flag And Similarity Summary
FAQ
Does Turnitin flag everything written by AI?
No. Turnitin's AI detection model is designed to be conservative and reports results with confidence thresholds. It flags text only when the model has high confidence that the content was AI-generated. This means some AI-generated text may go undetected, and conversely, some human-written text may be flagged incorrectly (false positives), though the latter is minimized through academic writing training data [1].
What does a *% score mean on Turnitin?
A *% (asterisk) score means Turnitin detected AI-generated content in the submission, but the overall percentage was below 20%. Instead of displaying a specific number like 5% or 12%, Turnitin uses the asterisk to indicate low-confidence detection. This does not mean the paper is free of AI content — it simply means the detected portion fell below the reporting threshold [3].
Can my instructor see which sentences were flagged as AI-written?
Yes. The Turnitin AI writing report includes sentence-level highlights that show exactly which portions of the text the model identified as likely AI-generated. Instructors can review these highlights alongside the overall percentage to understand the distribution of flagged content throughout the paper [2].
Is it possible to get a 0% AI score on Turnitin?
Yes. A 0% score means the detection model found no evidence that any portion of the submission was written by an AI tool. This is the only precise low numeric score Turnitin reports — all other sub-20% results are displayed as *% [3].
How can I check my Turnitin AI score before submitting to my university?
You can use a third-party service like Turnitin0.com, which provides genuine Turnitin AI and similarity reports before you submit to your institution. Simply upload your.docx,.pdf, or.txt file, and you'll receive the same AI score and flagged highlights that your instructor would see, typically within 10 minutes [4].