Turnitin Submission Check Process Explained

Published at June 04, 2026·By James Peng

Table of Contents

• What the Turnitin Submission Check Process Actually Is
• Step 1: Upload — What Happens When You Submit Your File
• Step 2: Processing — Status, Timing, and What “Still Loading” Means
• Step 3: Similarity Report — Matches, Percentages, and What They Do Not Mean
• Step 4: AI Writing Report — Highlights, *% , and 0%
• Step 5: Instructor Review — What Happens After Reports Generate
• What You Should Do Before the Official Submission Check
• FAQ
• Sources
• Related articles

What the Turnitin Submission Check Process Actually Is

When your course uses Turnitin, submission check means Turnitin compares your uploaded file against its databases and models, then returns reports your instructor can open inside the same assignment. Students often treat it as one “Turnitin score,” but Turnitin runs separate analyses:

Important boundary: A submission check is feedback for review, not an automatic misconduct verdict. Turnitin’s own guidance treats similarity percentages as a starting point for human review, and AI indicators as signals that require instructor judgment and institutional policy—not sole proof of cheating (Turnitin Guides: Using the AI Writing Report).

Why beginners confuse the stages: The LMS often shows one progress bar, then two different report icons. Similarity and AI use different models, different highlights, and different headline numbers. Reading them in the wrong order—panicking over AI before fixing a missing citation—is one of the most common mistakes during deadline week.

Scope note: This article covers standard coursework uploads through institutional Turnitin integrations. Exam proctoring, group-authorship disputes, and journal submission workflows follow different rules.

Step 1: Upload — What Happens When You Submit Your File

Upload is the handoff from your learning management system to Turnitin. You are not “logging into Turnitin” as a standalone app in most courses—you submit inside the assignment your instructor created.

What you do at upload

1. Open the correct course and assignment slot (draft vs final matters when resubmissions are limited).
2. Confirm the required file type—commonly .docx, .pdf, .txt, or .rtf depending on syllabus rules.
3. Attach the exact file you intend to be graded, not an older export with a similar title.
4. Click submit and wait for a confirmation screen or email receipt if your school sends one.

What Turnitin receives

Turnitin ingests the file text (and layout metadata where relevant), records the submission timestamp, and queues analysis. If your assignment allows multiple attempts, each new upload may generate a new report version; some courses treat the latest attempt as official, others lock after the first upload—check your syllabus.

Upload mistakes that skew later reports

• Wrong format: Previewing a Word file but submitting a PDF with different pagination can change match boundaries.
• Wrong document: A cover sheet from last semester merged into page one can spike similarity overnight.
• Too early: Uploading a half-finished draft “just to see” burns an attempt when resubmissions are disabled.
• Excluded content surprises: Instructors can exclude bibliography or quoted material from the similarity score; students sometimes compare their raw number to a classmate’s filtered view.

Worked scenario (composite): Sam uploads Essay_Final.docx to a Canvas assignment at 10:47 p.m. The page shows Submitted immediately, then Processing within a minute. Sam assumes the similarity percentage is final—it is not. Processing must finish before either report is trustworthy.

If you want to see how upload and report timing behave on your near-final file—not a generic demo—preview Turnitin reports while you still have a full revision window.

Preview your Turnitin reports before you submit →

Step 2: Processing — Status, Timing, and What “Still Loading” Means

Processing is Turnitin’s analysis queue. During this window, headline scores may be missing, grayed out, or marked pending.

Typical timing

• Short essays (under ~3,000 words): Often a few minutes; busy periods at term end can stretch longer.
• Long papers and theses: Can take noticeably longer because more text is indexed and matched.
• Peak deadline nights: Campus-wide submission spikes slow everyone’s queue equally—refreshing every ten seconds does not speed the server.

Turnitin’s help materials describe student-facing status labels (submitted, processing, complete) inside integrated assignments; exact wording varies slightly by LMS skin (Turnitin Help Center).

What to do while you wait

• Do not resubmit a duplicate file unless the assignment allows multiple attempts and you intend to replace the attempt.
• Note the timestamp in case you need to email your instructor about a late queue.
• Prepare your review plan: Similarity first when you used many sources; AI second when similarity is already clean but you machine-smoothed prose.

When processing seems “stuck”

If processing exceeds your course’s usual window (ask upper-year students or your writing center what “normal” is at your school):

1. Log out and back in—LMS sessions sometimes cache an old status.
2. Confirm you uploaded to the graded assignment, not a practice shell.
3. Contact your instructor with course, assignment name, and submission time—IT issues are common on the last night of term.

Processing completes when both enabled reports are ready to open (similarity and/or AI, depending on institutional settings).

Step 3: Similarity Report — Matches, Percentages, and What They Do Not Mean

After processing, Turnitin opens the Similarity Report (sometimes labeled Originality or plagiarism check in student-facing UI). The headline similarity score is a percentage of your submission that matches text in Turnitin’s search sets and your instructor’s comparison settings.

How to read the similarity report like an editor

1. Open source links, not only the percentage. Matched strings may be properly quoted, common phrases, or your own prior draft stored in the database.
2. Check filters. Instructors may exclude bibliography, quotes, or small matches. Your 18% and a classmate’s 9% are not comparable without knowing exclusions.
3. Prioritize uncited overlap. Missing quotation marks, page numbers, or paraphrase too close to a web summary are fixable before the instructor reads the same links.
4. Separate similarity from AI. Turnitin documents that the AI writing percentage is independent of the similarity score; AI highlights do not appear inside the Similarity Report view (Turnitin Guides: Using the AI Writing Report).

What similarity does not prove

• A high score is not automatic plagiarism—it is a triage signal.
• A low score is not automatic innocence—hidden uncited paraphrase can still fail policy.
• Matching your own prior work can appear if an earlier draft was indexed.

For research papers, similarity often clusters in literature reviews, block quotes, and reference-heavy sections. Beginners who treat the headline number as pass/fail skip the sentence-level work instructors actually perform.

Common beginner mistake: Rewriting the entire introduction because the similarity score “feels high,” when three uncited sentences in paragraph four caused most matches. Always read the color-coded strings before major surgery.

Step 4: AI Writing Report — Highlights, *% , and 0%

When your institution enables AI writing detection, Turnitin generates a separate AI writing report after the same upload finishes processing. This report highlights qualifying prose segments the model flags as likely AI-generated. It does not replace the similarity report, and it does not run the same way on poetry, scripts, code blocks, bullet lists, or tables (Turnitin Guides: Using the AI Writing Report).

Headline indicator vs highlighted sentences

Beginners fixate on the headline percentage. Instructors often read which sentences highlight and whether those sections match your disclosure statements. A short essay with a low headline number can still show dense highlights in the introduction if that section was heavily machine-polished.

Turnitin requires at least 300 words of qualifying prose in long-form writing for the AI report to run meaningfully, with an upper bound around 30,000 words in supported languages (Turnitin Guides: Using the AI Writing Report). Bullet-heavy outlines or code appendices may not contribute equally—another reason the headline number and highlights can feel out of sync.

How to read *% and 0% on the AI report

When you open the AI writing report, scores below 20% display as *% (an asterisk), not as single-digit percentages like “4%” or “11%.” Turnitin uses that display because false positives are more common in that range. 0% is the usual explicit low numeric outcome students screenshot.

Practical reading rule:

• 0% means the model did not flag qualifying prose at the headline level—but still open highlights if your syllabus requires disclosure of AI assistance.
• *% means “below 20%, shown cautiously”—not a free pass. Read highlighted sentences and compare them to your course AI policy.
• Higher explicit percentages mean more qualifying prose flagged; instructors are expected to apply human judgment, not auto-penalize from the number alone (Turnitin Guides: Using the AI Writing Report).

What the AI report does not do

• It does not prove you “used ChatGPT” by brand name—it models writing patterns.
• It does not guarantee the same headline on resubmission after edits.
• It should not be the sole basis for misconduct findings; Turnitin positions it as one signal alongside instructor review.

Order-of-operations tip: If similarity and AI both look concerning, fix uncited similarity clusters first. Citation work often shrinks AI highlights that sat on pasted summaries you later paraphrased properly.

Step 5: Instructor Review — What Happens After Reports Generate

Reports are inputs; instructor review is the decision stage. Policies differ by department, but the workflow usually looks like this:

1. Open the submission in the LMS grader view.
2. Read similarity matches with institutional filters applied (quotes, bibliography, small matches).
3. Open the AI writing report if enabled—review highlights, not only *% or 0%.
4. Compare to syllabus rules on generative AI, group work, and citation style.
5. Apply human judgment—ask for drafts, hold a conference, refer to integrity office, or grade normally.

Turnitin’s documentation stresses that AI detection should not be the only evidence in an academic misconduct case (Turnitin Guides: Using the AI Writing Report). Instructors may also weigh your revision history, in-class writing samples, and required AI disclosures.

What students see vs what instructors see

Some courses hide AI scores from students while instructors still review them—check your portal before you assume “no number means no flag.”

When to email before the graded decision locks

• High similarity you cannot explain (wrong file, template left in).
• AI highlights on sections you believed were fully human-written.
• Syllabus requires AI disclosure and you are unsure how to phrase it.

Short template:

Subject: [Course] [Assignment] – question after Turnitin reports
Hi Professor [Name], my submission shows [similarity/AI highlights] in [section]. I plan to [cite/rewrite/disclose]. Is that approach acceptable? Thank you, [Name]

What You Should Do Before the Official Submission Check

Use this checklist on the same file you plan to upload, while the assignment is still open for edits:

1. Confirm assignment slot — draft vs final, attempt limits, and due time zone.
2. Match file type — preview and submit the same export (.docx or .pdf).
3. Finish citations first — reference list and in-text citations in place before you interpret scores.
4. Run both reports when available — similarity triage, then AI highlights.
5. Fix similarity clusters before voice-polishing — uncited overlap often drives both reports.
6. Rewrite flagged prose in your own analysis — add course-specific examples instructors recognize.
7. Add required AI disclosures — use syllabus wording, not forum rumors.
8. Re-export and re-check after major edits if policy allows another run.
9. Keep dated drafts — useful if your instructor asks how you revised.
10. Avoid public screenshots — many programs treat reports as confidential.

Before you upload

Step 4 is where many students catch problems early: preview both similarity and AI on the file you plan to submit. If you have not done that yet, run your draft once while you can still edit citations and flagged sections.

Check your draft for similarity and AI detection →

FAQ

How long does Turnitin take after I submit?

Often a few minutes for typical essays, longer for large files or peak deadline traffic. Status stays processing until reports finish; exact labels depend on your LMS integration (Turnitin Help Center).

Can I see Turnitin reports before my instructor?

Sometimes. Many courses let students open the similarity report after processing; AI visibility varies by institutional settings. If your portal hides AI scores, ask your instructor what they will review—you may still need to fix flagged sections they can see.

Is similarity the same as the AI score?

No. Turnitin runs independent analyses. Similarity measures text overlap with sources; AI writing detection flags qualifying prose segments. Fix citations first when uncited overlap is present (Turnitin Guides: Using the AI Writing Report).

What does *% mean on the Turnitin AI report?

*% means the headline AI indicator is below 20%, displayed with an asterisk instead of a single-digit percentage because false positives are more common in that range. 0% is the explicit low number students usually screenshot. Always read highlighted sentences, not only the headline.

Does a low or *% AI score mean I am safe?

No automatic “safe” label exists. Instructors apply syllabus policy and human review. Low indicators still can include sentence highlights; high indicators still require context. Turnitin warns against using AI scores as the sole misconduct evidence (Turnitin Guides: Using the AI Writing Report).

What do instructors see that students do not?

Instructors typically see similarity matches, filter settings, AI highlights when enabled, and grading tools inside the LMS. Some schools hide AI headline scores from students while instructors retain full reports—confirm locally.

Can I check my paper before the official LMS submission?

Turnitin’s help center notes students generally cannot self-check inside Turnitin without submitting to an instructor-created assignment unless the school enables Draft Coach or practice slots (Turnitin Help Center). When policy allows, independent pre-submission services return official Turnitin similarity and AI writing reports on your own draft—confirm your handbook first.

Where can I preview both Turnitin reports on my own file?

Turnitin0 accepts .docx, .pdf, or .txt uploads and delivers official Turnitin similarity and AI writing reports for private review; it does not archive submitted papers into third-party databases. Confirm your syllabus allows outside checks before uploading coursework.

Will my preview match my LMS submission exactly?

Not always. File edits, export format, database updates, and instructor exclusion settings can shift results. Previews reduce surprises; they do not guarantee identical numbers on the graded attempt.

Should I fix similarity or AI first?

Similarity first when uncited quotes or pasted summaries drive matches. AI first when similarity is clean but highlights cover sections you machine-smoothed. Re-check both after major rewrites.

Sources

• Turnitin Guides: Accessing the Similarity Report
• Turnitin Guides: Using the AI Writing Report
• Turnitin Help Center: Can students check a paper before submitting?

Related articles

• Tips for Dealing with Turnitin's Ai Detection: What Actually Helps Before You Submit
• STEM Lab Reports vs Humanities Essays: Different Turnitin Failure Modes—and Fixes
• Can Turnitin Detect Gemini 1.0?
• How to Avoid Turnitin Flagging the Ai Detector
• International Students: Turnitin, Paraphrasing, and Citation Mistakes That Trigger AI Scores