Turnitin in Canvas: What Students See vs What Instructors See in the AI Report

In Canvas, the Turnitin AI report student vs instructor view is genuinely asymmetric: instructors see the AI writing indicator and can open a full AI report, while students by default see only the similarity side — and often no AI information at all. That gap surprises people on both sides of the gradebook. Students imagine a public verdict stamped on their work; instructors sometimes assume students already saw the number they’re about to bring up. Neither is true, and knowing exactly who sees what makes the whole conversation less tense.

Key takeaways

• The AI writing indicator is instructor-side only in the default Turnitin configuration. Students don’t get a version of it.
• Whether students can see the *similarity* report at all is a separate, assignment-level setting the instructor controls.
• There’s no automatic notification when a paper is flagged for AI. If it matters, a human tells you.
• Instructors see a percentage plus a highlighted report; both are estimates, not evidence.
• The strongest student response to any flag is process documentation, not a counter-percentage.

The student vs instructor view, side by side

Start with what each role actually gets after a submission lands in a Turnitin-linked Canvas assignment.

The instructor’s view. Inside SpeedGrader or the Turnitin report viewer, an instructor typically sees two separate things: the similarity score, which has been around for two decades, and the newer AI writing indicator. Clicking the indicator opens Turnitin’s AI writing report, where passages the model considers likely AI-generated are highlighted. We’ve walked through how instructors read the AI flag in SpeedGrader in detail, but the short version is: it’s a percentage, a set of highlights, and whatever judgment the instructor brings.

The student’s view. After submitting, a student may see their similarity score and report — if the instructor enabled that visibility for the assignment. Many do, some don’t, and some reveal it only after the due date. What the student does *not* see, in the standard setup, is anything about AI. No percentage, no highlights, no banner. The AI indicator simply isn’t part of the student-facing report.

That’s the core asymmetry. It isn’t a bug or a secret; it’s how Turnitin designed the feature, and it shapes how these conversations start.

Why the AI score stays on the instructor’s side

Turnitin has been fairly explicit that the AI indicator is meant as a signal for educators, not a grade. A few reasons the company and most institutions keep it instructor-side:

False positives are real. Detectors work on statistical patterns — how predictable word choices are, how uniform sentence rhythm is. Human writing that is formal, carefully edited, or produced by non-native English speakers can trip those patterns. Research from Stanford-affiliated authors in 2023 found detectors flagged non-native English essays at dramatically higher rates than native ones. Handing every student a raw percentage would generate a lot of unearned panic.

The number invites gaming. If students could watch their AI score live, the obvious move would be to iterate against it — tweak, resubmit, tweak again — which turns a probabilistic signal into a target. Instructor-side placement pushes effort toward writing rather than score-chasing.

Context is the point. An experienced instructor reads the flag alongside the student’s past work, in-class writing, and draft history. The percentage alone was never supposed to stand on its own, and most institutional policies now say so out loud.

What actually happens when a paper gets flagged

There’s no siren. Here’s the typical sequence, from the instructor’s chair:

1. The instructor opens SpeedGrader and notices the AI indicator is elevated on one submission.
2. They click through to the AI report and look at which passages are highlighted — is it the whole essay, or three suspiciously smooth paragraphs?
3. They compare against context: the student’s earlier submissions, discussion posts, anything written in class.
4. If doubt remains, they email the student or raise it in office hours. This is usually the first moment the student learns any score exists.

Notice what’s absent: automatic penalties, automatic reports to an integrity office, automatic anything. Some institutions have workflows that kick in at that fourth step, but the trigger is human judgment, not the number itself.

What this means for you as the submitter

Don’t burn energy guessing your score. You can’t see Turnitin’s AI number, and reverse-engineering it from third-party tools is imprecise — different detectors disagree with each other constantly. If you want a sanity check on how statistically “smooth” your writing reads, you can check how your draft reads before submitting, but treat any tool’s output as a rough signal, not a prediction of Turnitin’s.

Write like a person with a history. Draft in a tool that keeps version history. Keep your outline, your notes, your half-finished paragraphs. If a flag ever surfaces, using version history to defend yourself is the single most effective move available — it shifts the conversation from a probability to a paper trail.

If it comes up, ask specific questions. Which sections were highlighted? Is the concern the whole document or part of it? What does the instructor’s policy say the flag means? Specific questions signal good faith and move the discussion onto ground where your evidence counts.

Know your school’s policy. Many institutions explicitly prohibit acting on the AI indicator alone. Finding that sentence in your academic integrity policy before you need it is ten minutes well spent.

A quick scenario

Priya submits a literature review to a Canvas assignment. She sees her similarity report — 8%, mostly citations — and moves on with her week. Her instructor, meanwhile, sees an AI indicator in the 40s and highlights across the methods summary. He checks Priya’s discussion posts, which read similarly formal, then emails her anyway to ask about her process. Priya sends her Google Docs version history showing eleven days of edits.

The whole exchange takes two emails, because both sides understood what the report does and doesn’t prove. That’s the system working — and it started from an asymmetry Priya couldn’t see.

Frequently Asked Questions

Can students see the Turnitin AI score in Canvas? By default, no. Turnitin’s AI writing indicator is shown to instructors in the report view, and it is not included in the student-facing similarity report unless an institution or instructor deliberately shares it. Most students only learn a score exists when an instructor brings it up, which is why keeping drafts and version history matters more than trying to guess your number.

Do students and instructors see the same similarity score? Usually yes, if the instructor has chosen to let students view the similarity report at all. That visibility is an assignment-level setting. The AI writing indicator is different: it sits on the instructor’s side of the report and has no student-facing equivalent in the default configuration.

Why does Turnitin hide the AI score from students? Turnitin has said the indicator is meant as a signal for educators to interpret with context, not a verdict to hand to students. A raw percentage with no explanation invites panic, endless self-testing, and attempts to game the number. Keeping it instructor-side pushes the conversation toward process and evidence instead of a single statistic.

Will I be told if my paper gets flagged for AI? Only if your instructor decides to raise it. There is no automatic notification to students when the AI indicator is high. Institutions generally instruct faculty to treat the flag as a starting point for a conversation, so if it matters, you will hear about it from a person, not from Canvas.

What should I do if my instructor mentions a high AI score? Stay calm and lean on your process. Ask which sections were flagged, then share your version history, outlines, notes, and earlier drafts. Detectors are known to produce false positives, especially on formal or non-native English writing, and most schools bar using the score as sole evidence. Documentation of how you wrote is far more persuasive than arguing about the percentage.

The short version

Instructors see the AI indicator; students, by default, don’t. Nothing about that asymmetry should change how you write — it should only change how you prepare. Keep your drafts, know your policy, and if you’re curious how your writing reads to a detector before the stakes are real, run it through a checker yourself. You can browse more guides on the blog for the rest of the Canvas-and-Turnitin picture, or see how our plans work if you want more than a quick check.