Turnitin Clarity Explained: What the New Policy Tiers Mean for Your Paper

A grad student we work with submitted a final paper through her university's Turnitin Clarity portal last semester. The AI score came back at 76%. Two years ago, that number would have triggered an academic integrity meeting and possibly a hearing. This time, her instructor opened the Clarity timeline, saw 14 hours of drafting spread over six sessions across two weeks, and closed the case before it started.

That's the shift Turnitin is trying to engineer. Clarity isn't a better AI detector. It's a different model for evaluating student work — one where the writing process is the evidence, not the final text. Whether your institution has adopted it, is piloting it, or hasn't decided, the policy tier framework Turnitin is pushing alongside Clarity will probably affect how your work is evaluated in the next year or two. This guide explains what Clarity actually is, the three policy tiers institutions are now choosing between, how the writing-process recording works in practice, what it can and can't prove, and how to use it without anxiety.

The 90-second version of what Clarity is

Turnitin Clarity is a writing tool that records your drafting process and produces a timeline showing how the document evolved. It's installed as a Microsoft Word add-in or accessed through a web editor. When you draft in it, every keystroke, paste, and session boundary is captured. When you submit, your instructor sees both the final paper and the timeline of how you wrote it.

The pitch is straightforward: AI-generated text typically appears in a document as large pasted blocks with minimal subsequent editing. Human writing typically shows incremental construction — adding sentences, deleting them, restructuring paragraphs, returning to earlier sections days later. The timeline makes that distinction visible.

Turnitin launched Clarity in late 2024 as a paid add-on to its standard plagiarism and AI-detection product. Adoption accelerated through 2025 as institutions looked for responses to the false-positive problem in AI detection. By mid-2026, several large university systems have made Clarity available to instructors, and a handful have begun requiring it for high-stakes assignments.

Clarity is not free for institutions. It's a per-seat cost layered on existing Turnitin subscriptions. Whether your institution has it depends on what your administration has chosen to fund.

Why this changed in 2026

A few converging pressures pushed institutions toward something like Clarity.

The first was the false-positive problem becoming legally costly. The Newby v. ECU settlement in early 2026 established that institutions can face real liability for academic integrity actions based primarily on AI-detection scores. Administrators who had previously treated detector scores as definitive started looking for backup evidence — and process recording was the obvious answer.

The second was peer-reviewed research consistently showing detector limitations. Multiple 2024-2025 studies documented false-positive rates significantly higher than vendor claims, with particularly elevated rates for non-native English speakers, formal academic writing, and edited writing. The cumulative weight of these studies made the "we trust the detector" stance harder to defend.

The third was instructor exhaustion with adjudicating contested cases. Every flagged paper turns into a meeting, a back-and-forth, sometimes a formal hearing. Instructors started asking for tools that produced fewer ambiguous cases in the first place. Process recording, whatever its limitations, produces clearer evidence than statistical patterns alone.

The three policy tiers

Alongside Clarity, Turnitin has been promoting a policy tier framework that institutions and instructors are now using to set expectations on a per-course or per-assignment basis. The three tiers, as commonly implemented:

Tier 1 — No AI Permitted. Students may not use AI tools in any capacity for the assignment. Grammar checkers and spell-check are typically excluded from the prohibition (institutional language varies). Clarity is often required as the verification mechanism — students draft in Clarity, the timeline serves as evidence of unassisted authorship.

Tier 2 — AI Permitted for Outlines and Drafting Support. Students may use AI for brainstorming, outlining, and feedback on drafts but must write the final text themselves. Disclosure of which tools were used is required. Clarity may be used to verify that the final text was constructed rather than pasted.

Tier 3 — AI Permitted with Disclosure. Students may use AI broadly, including for generating text, as long as they disclose what they did and take responsibility for the content. This tier is more common in upper-level courses, graduate work, and professional programs where the assessment criteria focus on the student's judgment and synthesis rather than the act of writing.

Some institutions have added a fourth tier sometimes labeled "AI Required" — courses or assignments that explicitly teach AI-assisted writing as a skill. This is rare but growing.

The shift this framework drives is meaningful: instead of one institution-wide policy that pretends AI doesn't exist, courses and assignments declare their own expectations. Students need to check the policy for each assignment, not assume the default. Instructors need to be explicit, not implicit.

How Clarity actually works

The technical reality is more granular than the marketing.

When you install the Clarity Word add-in, the tool starts recording your work on any document you mark as a Clarity-enabled assignment. The recording captures:

• Every keystroke, with timestamps
• Every paste, with the length of pasted content flagged
• Every cut, copy, and undo
• Session boundaries (when you opened, closed, or saved the document)
• Time-active vs idle (the tool distinguishes between you reading and you typing)

When you submit, the timeline is rendered as a visualization for your instructor. They can scrub through it, see major pasted blocks highlighted, and check the overall pattern of your drafting.

The web-editor version works similarly but captures fewer dimensions — keystrokes and pastes are recorded; session-level behavior is captured at a coarser granularity. Most institutions that require Clarity prefer the Word add-in because the data is richer.

The data is stored by Turnitin, accessible to your instructor for the duration of the course, and subject to your institution's data retention policy. It's not shared with other instructors or with the general detector training data.

What Clarity can and can't prove

Clarity is good evidence of process. It is not perfect evidence of authorship.

What it shows clearly:

• That a document was constructed over time rather than pasted in
• That the author was actively typing during sessions
• That major restructurings happened (insertions, deletions, paragraph reorganizations)
• That writing happened across multiple sessions on different days

What it doesn't show:

• Whether the typed content was original or copied from another source. A student who reads an AI-generated essay on one screen and types it into Clarity on another produces a timeline that looks like authentic drafting. Clarity cannot distinguish typing from transcribing.
• Whether ideas, arguments, or structure came from AI. The intellectual contribution is invisible to a process recording.
• Whether a collaborator did substantial work. Two students working together produce a timeline that looks the same as one student working alone.

This is why institutions adopting Clarity usually pair it with other evidence — assignment design that's hard to AI-generate, in-class writing components, conversation-based assessments. Clarity is one input, not a verdict.

For students: how to use it without anxiety

If your institution or instructor requires Clarity, the practical guidance is straightforward.

Draft in the tool. Don't draft elsewhere and paste in. The whole value of Clarity is the recording, and pasting in large blocks creates exactly the pattern instructors are trained to flag. Even if your pasted text is your own, the timeline can't tell.

Take natural breaks. Real drafting happens across multiple sessions. A timeline showing one continuous 4-hour writing session at 11 p.m. before the deadline doesn't look great, even if it's genuine. If you're capable of working in 45-90 minute chunks across multiple days, do that.

Use the editor for ideation, not just transcription. If you're brainstorming, do it inside the tool. If you're planning your outline, type it as bullet points first. The richer your drafting trail, the more clearly your process shows.

If you used AI tools and your assignment allows it, disclose them. The disclosure is what protects you. Clarity will show pasted blocks if you used AI text; the disclosure explains those blocks in context. Hidden AI use plus visible paste patterns is the worst combination for your case.

Don't try to "game" the timeline. Some students have tried things like typing out AI-generated text character by character to fool Clarity. Instructors using Clarity often look for unnatural typing patterns — perfectly even keystroke timing, no backspacing, no rephrasing. Authentic drafting includes false starts and revisions. Trying to fake those well enough to fool a trained reader is more work than just doing the assignment.

For instructors: how to set policy

A few practices we've seen work for instructors implementing Clarity at the course level.

State the policy explicitly in the syllabus. Don't assume students know which tier applies to your course. Write it out: "This course operates under Tier X. AI tools may [or may not] be used for [specific purposes]. Clarity timelines [are/are not] required for submitted work."

Distinguish between assignment types. Lower-stakes assignments (weekly journals, in-class writing) may not need Clarity. Higher-stakes assignments (final papers, capstone projects) may. Telling students which is which removes ambiguity.

Pair Clarity with assignment design. The strongest defense against AI-generated submissions isn't detection; it's assignment design that requires specific, contextual, original thinking. Clarity adds process evidence on top of well-designed assignments.

Use Clarity as evidence, not verdict. A flagged timeline is a reason to talk to the student, not a reason to issue an integrity notice. The conversation reveals what the timeline alone cannot — whether the student understands the work, whether they can explain their decisions, whether they can extend their argument.

Train graders consistently. If you have TAs reviewing Clarity timelines, they need shared training on what to flag. Inconsistent flagging across graders is a fairness issue students will notice.

How Clarity interacts with detection

Many institutions are running Clarity alongside Turnitin's similarity check and AI detection rather than replacing them. The combined output gives instructors three signals:

• Similarity score (how much overlaps with other sources)
• AI score (how much matches AI-generation patterns)
• Process timeline (how the document was drafted)

When these three signals agree, the case is usually clear. When they disagree, the case requires human judgment. A high AI score with a robust process timeline is the most common ambiguous case — and is usually resolved in the student's favor when other evidence (specific knowledge of the topic, ability to discuss the work, consistent quality across other assignments) supports it.

For students who've been falsely flagged before any of this existed, our guide to appealing a false AI-detection flag covers the playbook. For students wondering whether to disclose AI use that was permitted, our AI-use disclosure guide covers the templates. And for students whose courses sit in Tier 2 or Tier 3 — where editing tools are explicitly allowed with disclosure — our AI proofreader produces the kind of tracked-changes record that fits naturally alongside a Clarity timeline.

See the Full AI Proofreader

Tracked-changes editing with version history. Free tier includes every feature.

Frequently asked questions

Q: Does my university have Turnitin Clarity?

Check with your registrar, library, or academic affairs office. Clarity is a paid add-on to Turnitin's standard product, so not every institution that has Turnitin has Clarity. Many large university systems acquired it through 2025; many smaller institutions are still evaluating. Even at institutions that have Clarity, individual instructors decide whether to require it on their assignments. The syllabus is the most reliable source for whether you need to use it in a specific course.

Q: Is Clarity better than the AI detector at proving I didn't use AI?

Better in some ways, not in others. Clarity provides process evidence that's harder to dispute than a probability score; instructors generally trust a robust drafting timeline more than they trust an 80% AI flag with no other evidence. But Clarity can't distinguish authentic typing from transcribing, can't show intellectual contribution, and can't prove what you didn't do — it can only show what your typing patterns looked like. The strongest defense remains the combination of process evidence plus the ability to explain and extend your work in conversation.

Q: What if I draft in Google Docs and paste into Clarity at the end?

This is the worst-case workflow. Your final paste shows up as a large pasted block, which instructors are trained to flag. Your previous drafting history is invisible to Clarity. Even if your work is entirely your own, the timeline looks like an AI-paste pattern. If your assignment requires Clarity, draft in the Clarity tool from the start. If you prefer Google Docs, ask your instructor whether the Google Docs version history (which is also a process record) satisfies the requirement.

Q: Can my instructor see what I deleted and rewrote in Clarity?

Yes. The full timeline includes deletions, restructurings, and revisions. This is generally good for you — it shows authentic drafting behavior, including the false starts and corrections that real writing involves. It does mean that early drafts you regretted or strong opinions you toned down are visible to your instructor as part of the process record. Don't write anything in the Clarity tool you wouldn't be comfortable having your instructor see. If you want to brainstorm freely first, do it elsewhere (in a notebook, a non-Clarity document) and bring the polished ideas into Clarity for the actual drafting.