Turnitin Feedback Studio in Canvas: Navigating the AI Indicator and the Layers Panel

In Feedback Studio, the layers panel controls what overlays the essay — grading marks, similarity highlights, grammar feedback — while the AI indicator sits apart in the side panel as a small percentage badge that opens a separate AI writing report in a new tab. The Turnitin Feedback Studio AI indicator is probably the most-missed element in the whole interface: instructors toggle every layer looking for it, and it was never a layer to begin with. Once you see how the viewer is actually organized — overlays for checkable feedback, a separate report for the probabilistic stuff — the layout stops fighting you.

Key takeaways

• Layers = toggleable overlays on the document: grading (QuickMarks, comments), similarity (source-match highlights), and grammar feedback where enabled.
• The AI indicator is not a layer. It’s a percentage badge in the side panel; its highlights live in a separate report that opens in a new tab.
• Similarity highlights and AI-report highlights assert opposite things: “this text exists elsewhere” vs “this text reads machine-made.”
• Students never see the AI element — instructor and admin eyes only, per Turnitin policy.
• A blank badge usually means the submission didn’t meet the AI analysis minimum, not that the check passed.

Getting there from Canvas

The path is short: open the submission in SpeedGrader and launch the Turnitin report — depending on your institution’s configuration, from the similarity score link or the assignment’s Turnitin panel. Feedback Studio opens with the student’s document center-stage and a toolbar of icons down the side.

That side toolbar is the whole game. Everything Feedback Studio can tell you about this submission is reachable from it, and the two things this article cares about — the layers panel and the AI indicator — are both there, doing very different jobs. (For what’s visible *before* you launch, from Canvas’s own side, see how the Turnitin AI flag appears from SpeedGrader’s side.)

The layers panel: what overlays the document

Feedback Studio can paint several kinds of information directly onto the essay, and layers are how you control the traffic:

• Grading layer — your QuickMarks, inline comments, and feedback annotations.
• Similarity layer — the colored highlights tying passages to matched sources, with the source breakdown in the panel.
• Grammar feedback layer — automated writing feedback, where institutions have it enabled.

The active layers control lets you flip each independently, and the workflow reason is obvious after one crowded essay: grading with match highlights everywhere is unreadable, and reviewing matches under a blanket of your own comments is worse. Toggle to what the current task needs — comment with similarity off, verify sources with grading off.

Notice what all layer content has in common: it’s *positional* feedback about specific passages, and it’s either something you wrote or something you can check. A similarity highlight names its source; you can open the source and judge the match yourself. That property — checkability — is quietly the organizing principle of the interface, because the one analysis that lacks it lives somewhere else entirely.

The AI indicator: the badge that isn’t a layer

Look below the similarity tools in the side panel and there it is: a small percentage badge — the AI writing indicator. Not a toggle, not an overlay, no highlights appearing on the document in front of you. Click it and the AI writing report opens in a new browser tab: the submission’s analyzed prose with highlighted segments the classifier reads as likely machine-generated, and in the current report design, distinct marking for text that looks AI-generated versus AI-generated-then-paraphrased.

Three behaviors worth internalizing:

Blank doesn’t mean clean. The badge shows no percentage when the analysis didn’t run — most often because the submission had too little qualifying prose (a few hundred words minimum, after quotes, reference lists, and non-prose content are stripped). Absence of a number is absence of a measurement.

Low numbers carry an asterisk-grade caveat. Turnitin explicitly flags lower-range scores as less reliable. The report’s own framing keeps repeating: probability, not proof.

It’s instructor-only. Students never see the badge, in Canvas or anywhere else. Their view of Feedback Studio simply doesn’t include the element.

We’ve unpacked the report’s internals — what the percentage covers and how to read the highlighting — in reading the AI writing report’s percentages and highlights.

Why the separation is the right mental model

It’s tempting to read the interface split as clutter management. It’s better read as epistemology.

Similarity highlights overlay the document because they invite verification: here’s the passage, here’s the source, decide. AI highlights open in a separate report because there is nothing to verify — a classifier’s confidence about statistical texture, sourceless by nature. The same sentence rarely carries both marks, since matched text is generally human-written and generated text generally matches nothing. The two analyses even fail differently: similarity fails by over-flagging quotes and boilerplate, which you resolve by looking; the AI model fails by misreading polished human prose — formal register, non-native English writers — which you *cannot* resolve by looking harder at the report.

So the interface is telling you how to treat each signal. Layer content: adjudicate from the evidence in front of you. Side-panel badge: investigate — version history, drafts, the student’s known voice, a conversation. An integrity decision built from the badge alone runs directly against the vendor’s own guidance, and honestly, against the interface design itself.

Practical notes for both sides of the desk

Instructors: learn the badge’s location once and stop hunting layers for it. Use layer toggles aggressively — one task, one layer set. And when the AI percentage is the concern, move the investigation *out* of Feedback Studio: the report can’t answer process questions, and process questions are the ones that matter.

Students: nothing in your Feedback Studio view shows AI analysis, so don’t read anything into what you can or can’t see there. What you control is upstream: writing where version history accumulates, keeping drafts, and knowing your own statistical profile before anyone else measures it — you can see your own draft’s heatmap first and find out whether your natural register reads machine-smooth *before* it surprises you on someone else’s screen. Where the fair-use line sits for that kind of self-checking is spelled out with our pricing, and the rest of the ecosystem — missing scores, file-type effects, LMS differences — is covered across more Turnitin guides.

One viewer, two kinds of claims: overlays you can check, and a badge you can only weigh. Navigate it that way and Feedback Studio makes considerably more sense than its icon toolbar suggests.

Frequently asked questions

Where exactly is the AI indicator in Feedback Studio?

In the side panel of the viewer, below the similarity tools — a small percentage badge rather than a headline number, which is why so many instructors miss it for months. Clicking it opens the full AI writing report in a new browser tab, with the analyzed text highlighted passage by passage. It is not one of the toggleable layers, it never overlays the document view the way similarity highlights do, and if the submission fell below the minimum prose threshold, the badge shows no percentage at all.

What are the layers in Feedback Studio?

Layers are overlay sets you can toggle on the document view: the grading layer with QuickMarks and inline comments, the similarity layer with its colored source-match highlights, and, where institutions enable it, a grammar-feedback layer. The active layers panel lets you switch each on or off so you can grade without match-highlight clutter or review matches without comment clutter. The AI analysis is deliberately not in this system — it renders in its own report, in its own tab.

Why do the AI highlights open in a separate tab instead of overlaying the essay?

Because the two analyses make different kinds of claims. A similarity highlight points at checkable evidence — this passage, that source — so overlaying it on the document invites verification. The AI report’s highlights mark passages a classifier finds statistically machine-like, with no source to check, and Turnitin presents that probabilistic judgment in a separate report with its own framing and caveats. Keeping it out of the grading surface is a quiet design argument: this signal deserves a different, slower kind of attention.

Do the similarity highlights and AI highlights mean the same thing?

No — they’re opposite claims that look confusingly similar. A similarity highlight says this text already exists somewhere and names the source. An AI-report highlight says this text is probably novel but reads machine-generated. The same sentence rarely triggers both, since matched text is usually human-written and generated text usually matches nothing. Mixing them up leads to real misreadings, like citing a similarity match as evidence of AI use — two different instruments, two different failure modes, two different conversations with the student.

Can students open the AI writing report on their own submission?

No. Students with report access can view their similarity results, but the AI indicator and its report are restricted to instructors and administrators — Turnitin policy, in every LMS. A student’s Feedback Studio view simply has no AI element in it. If you’re a student trying to anticipate that instructor-side number, your only option is external: check your draft with detection tools yourself before submitting, and keep your version history as evidence that the writing process was yours.