How Blackboard Ultra Surfaces the Turnitin AI Score Differently From Original Blackboard

In Blackboard Ultra, the Turnitin AI score only exists inside the Turnitin viewer you launch from a submission — Ultra’s gradebook shows no Turnitin columns — whereas Original Blackboard’s Grade Center at least surfaced similarity icons inline before you clicked through. The Blackboard Ultra Turnitin AI score experience is the same detection engine wearing a different, more buttoned-up interface, and instructors migrating from Original keep tripping over the same question: *where did everything go?* Short answer: behind the LTI launch. Here’s the longer one.

Key takeaways

• The AI writing indicator has never been a Blackboard column in either course view — it lives inside Turnitin’s viewer, instructor-only.
• What changed with Ultra is the similarity glance-view: Original’s Grade Center showed inline Turnitin score icons; Ultra’s gradebook shows grades, and the reports sit behind a launch into Turnitin.
• The move from the classic Building Block to LTI is what drove this — Turnitin now hosts its own interface rather than writing data into Blackboard’s.
• Detection itself is unchanged. Same engine, same thresholds, same instructor-only AI visibility, regardless of course view.
• SafeAssign is a separate Blackboard-native checker and not part of any of this.

Two course views, two integration generations

The Original/Ultra split isn’t just a visual refresh — the two course views typically talk to Turnitin through different plumbing, and the plumbing dictates what you see.

Original courses historically used Turnitin’s classic Building Block, an integration deep enough to write data back into Blackboard’s own interface. That’s why Grade Center could show a colored similarity icon per submission: the score was pushed into Blackboard’s tables, and clicking the icon opened the full report. Whatever you think of Original’s aesthetics, the at-a-glance density was genuinely useful — one screen, thirty submissions, thirty similarity scores.

Ultra connects through LTI, the modern standard Turnitin has moved all its LMS integrations toward as the old Building Block was deprecated. LTI is a cleaner contract: Blackboard sends the submission and receives a grade back, and everything else — the similarity report, the feedback layer, and the AI writing indicator — stays inside Turnitin’s hosted viewer. Blackboard no longer owns a copy of the score data to decorate its gradebook with.

The result is the trade every LTI migration makes: more consistency and easier maintenance for the institution, fewer inline conveniences for the person grading at 11pm.

Where the AI score actually is in Ultra

The path, as an instructor: open the course, open the assignment’s submissions, open a student’s submission, and launch the Turnitin report. You land in the Turnitin viewer. The similarity layer is front and center; the AI writing indicator is in the side panel — a percentage that, when clicked, opens the full AI writing report in a new tab, with the suspect passages highlighted.

Three things about that indicator that hold in Ultra exactly as everywhere else:

• It’s instructor-only. Students never see it, in any LMS, under any setting. That’s Turnitin policy.
• It has a minimum. Short submissions — under a few hundred words of qualifying prose — get no AI score at all, so blank isn’t a verdict.
• It’s a probability estimate. Turnitin’s own guidance says not to treat it as proof, and flags extra uncertainty on low-percentage results.

If you’re coming from classic Blackboard, the muscle-memory difference is just the entry point. We’ve covered how the AI indicator relates to Grade Center in classic Blackboard — in Original you started from a score you could already see; in Ultra you start from a gradebook that tells you nothing and click through on faith.

What migration does and doesn’t change

Institutions moving Original → Ultra generate a predictable wave of “Turnitin is broken” tickets each fall. Worth separating what actually changed from what didn’t:

Didn’t change: the analysis. The same submission produces the same similarity matching and the same AI estimate regardless of course view. Turnitin doesn’t know or care which Blackboard skin launched it. A paper that would have flagged in Original flags identically in Ultra.

Changed: the surface area. No inline score icons in the gradebook. Report access routed through the submission view. Assignment setup uses the LTI workflow rather than the old Turnitin assignment type, so options live in Turnitin’s settings panel instead of Blackboard’s assignment editor. And old courses copied forward from Original sometimes carry legacy Turnitin links that no longer launch cleanly — those need rebuilding as LTI assignments, which is an admin task, not a detection change.

For instructors, the honest adjustment is workflow speed: triaging a big class by similarity score now means opening the Turnitin side rather than scanning a Grade Center column. For students, nothing about your risk profile changed at all — the same text produces the same numbers.

Don’t confuse any of this with SafeAssign

Blackboard also ships its own originality checker, SafeAssign, which predates the AI era and is a completely separate product from Turnitin. An Ultra course might use SafeAssign, Turnitin, both on different assignments, or neither — that’s per-assignment instructor choice, and the reports look nothing alike. SafeAssign’s percentage is a text-matching measure; it is not an AI-writing estimate. We’ve written up whether SafeAssign detects AI at all separately, because conflating the two leads students to entirely wrong conclusions about what was checked.

Quick orientation: a report with source matches and a Blackboard-styled interface is SafeAssign; a launch into Turnitin’s viewer with the side-panel percentage is Feedback Studio, and that’s where an AI indicator might live.

The practical takeaways

Instructors: budget the extra click, and resist letting the friction change your standards — an AI percentage you had to dig for is exactly as uncertain as one that was handed to you inline. It’s one signal, it false-positives on real students, and it deserves the same due process in Ultra that it did in Original.

Students: the interface reshuffle changes nothing for you. Your similarity report still won’t include an AI number, and the score instructors see is produced the same way it always was. If you want to know how your writing reads to detection models before anyone institutional runs it, check your draft against a detector first — self-knowledge is the one part of this pipeline you control. How we frame the fair-use line on that kind of checking is on our pricing page, and there are more LMS guides covering the Canvas and Moodle versions of this same story.

Same engine, new furniture. Once you know where the one important number lives — inside the viewer, behind the launch, instructor-only — Ultra’s version of Turnitin holds no further mysteries.

Frequently asked questions

Where do I find the Turnitin AI score in Blackboard Ultra?

Open the assignment’s submissions, launch the Turnitin report for a student, and look in the side panel of the Turnitin viewer — the AI writing indicator sits there as a percentage, and clicking it opens the full AI report in a new tab. The score never appears in the Ultra gradebook itself. If you’re used to Original’s Grade Center showing inline Turnitin icons, the extra click into the viewer is the main adjustment: in Ultra, everything Turnitin-related lives on Turnitin’s side of the LTI launch.

Why doesn’t the Ultra gradebook show a Turnitin column like Grade Center did?

Because the integration generation changed. Original’s classic Building Block wrote similarity data into Grade Center, so you got colored score icons at a glance. Ultra connects through LTI, where Turnitin passes back a grade and hosts everything else — similarity report, AI indicator, feedback — inside its own viewer. The AI score was never a gradebook column in either version, but in Ultra even the similarity glance-view is reduced, which makes the whole report feel more hidden than it used to.

Did moving to Ultra change what the Turnitin AI detector analyzes?

No. The detection itself is identical — the same Turnitin engine processes the submission whether it arrived through Original, Ultra, Canvas, or Moodle, and the same thresholds and minimum word requirements apply. What changes between Original and Ultra is purely presentation: how many clicks sit between an instructor and the report, and what summary information Blackboard shows before the Turnitin viewer opens. A paper doesn’t score differently because your institution switched course views.

Can students see the AI score in either Blackboard version?

No. Turnitin shows its AI writing indicator to instructors and administrators only, and that policy is Turnitin’s, not Blackboard’s — so it holds in Original, Ultra, and every other LMS. Students who can view their similarity report will still find no AI percentage in it. If you’re a student trying to gauge how your writing reads to detection models, the only route is checking it yourself with external tools before submitting, because the institutional side will never display that number to you.

Is SafeAssign the same thing as Turnitin’s AI detection in Ultra?

No — SafeAssign is Blackboard’s own originality checker, a separate product that exists alongside any Turnitin integration a school licenses. Historically SafeAssign focused on text matching rather than AI detection, and any originality percentage it reports is a similarity measure, not an AI-writing estimate. If your course uses SafeAssign only, the Turnitin AI indicator doesn’t exist for your assignments at all. Which checker runs depends on what your instructor attached to each individual assignment.