Direct answer
An original content checker with an AI detector runs two separate analyses on the same document: a similarity check that compares your text against a database of published and previously submitted work, and an AI writing check that estimates how much of your prose appears machine-generated. The two numbers are independent — a draft can be 0% similar and still carry a high AI indicator, or be heavily matched to sources while showing no AI signal at all [1]. Reading them together, rather than as a single "originality score," is what tells you whether your draft is genuinely yours and how an instructor's report will look.
What Does an Original Content Checker With AI Detection Actually Measure?
The similarity half of the report measures overlap, not authorship. Turnitin compares your submission against its repository of student papers, its current and archived internet index, and its publication database, then reports the proportion of your text that matches sources in that database [2]. That percentage is a matching figure — it counts every string of words that appears elsewhere, whether or not you cited it, and whether or not you intended to copy it. A reference list, a block quotation, and a standard method description all inflate the number without any integrity problem existing.
The AI half measures something structurally different: statistical patterns in your sentence construction. The detector looks at predictability and uniformity across qualifying long-form prose and returns an estimate of how much of that text it attributes to AI generation [1]. It does not search a database of AI outputs to find a match — there is no "source" to point at, which is why the AI report shows a percentage band rather than a list of highlighted sources. Understanding this distinction is the single most useful thing you can do before you open a report, because students routinely misread an AI percentage as an accusation of copying.
The practical consequence is that originality is a judgement you make from evidence, not a number you receive. A 34% similarity score built entirely from correctly cited sources is a well-referenced paper; a 9% score that includes one unattributed paragraph lifted from a lecture slide is a real problem [2]. This is why the report shows every match with its source and location instead of just a total. If you want a checker that is actually useful, judge it on whether it gives you match-level detail and a separate AI indicator — not on whether it produces one reassuring headline figure.
Why Does Turnitin Show an Asterisk (*%) Instead of a Percentage for AI Writing?
An asterisk replaces the number when the detected AI writing falls below Turnitin's minimum confidence threshold, which is 20%. In that band the system is telling you it found some signal but not enough to report a reliable figure, so it declines to give one rather than present a precise-looking percentage it cannot stand behind [3]. A *% result is therefore a low-confidence signal, not a hidden high score and not a clean bill of health. Treat it as "below the reporting threshold" and nothing more.
The indicator also only analyses text that meets its length and language conditions — typically longer-form English prose — so short answers, tables, bullet lists, and reference sections are excluded from the calculation [3]. This has a counter-intuitive effect on how you read your own report. A 4,000-word essay with 1,200 words of quoted material may have its AI percentage calculated over a much smaller body of text than you expect, which is why the percentage can look disproportionate to the section you are worried about. Always check what was analysed before reacting to the number.
It is equally important to know what the indicator is not. Turnitin describes its AI writing detection as an estimate that supports academic judgement rather than a determination of misconduct, and it explicitly advises that the score should not be used as the sole basis for an integrity decision [3]. False positives occur on highly formulaic writing — literature review templates, methods sections, and non-native-speaker prose that has been heavily edited for consistency. If your report flags text you wrote yourself, the correct response is to keep your drafts, version history, and notes, because that evidence is what a human reviewer will actually weigh.
How Can Students Preview Both Their Similarity Report and AI Score Before Submitting?
Pre-submission checking is a revision tool, and its value depends entirely on timing. Reviewing a report a week before the deadline lets you paraphrase an over-matched paragraph, add a missing citation, or rewrite a section that reads too uniformly — all of which are ordinary editing decisions [4]. Reviewing it ten minutes before the deadline turns the same information into panic with no room to act on it. Build the check into your drafting schedule, not your submission ritual.
There is also an institutional caveat you should settle first. Policies on draft checking vary between universities and between individual modules: some institutions provide their own draft-submission routes, some permit external checkers, and some treat any pre-submission upload as a breach [4]. Check your module handbook or ask your tutor before you upload anything, and prefer a service that does not add your file to a student paper repository — otherwise a draft you check could later register as a match against your own final submission.
When you do review a report, work through it in a fixed order. Start with the AI indicator and note whether it is a number or an asterisk [3]. Then open the similarity report and sort matches by size, reading the largest ones first to confirm each is properly attributed [2]. Finally, compare the flagged passages against your own drafts and notes — if you can point to the version where that sentence first appeared, you have the evidence you need regardless of what the percentage says. That sequence turns a report from a verdict into a checklist.
Reading a report only helps if the report is the real thing. Turnitin0 produces the same AI detection and similarity PDFs your instructor sees in the LMS, delivered in under 15 minutes in 98% of cases, with your file kept out of the student paper database — so you can review the match detail and the AI indicator side by side, on your own draft, while there is still time to revise.
※ Turnitin0.com - Actual Turnitin AI Report Cover, Score, Flag And Similarity Summary
FAQ
Is a 0% similarity score the same as original content?
No. Zero similarity means nothing in your text matched the database Turnitin searched, which can happen with highly specific or very recent material as easily as with genuinely original writing. Originality is established by your sources and your own drafting record, not by a low match total [2].
Does an asterisk (*%) on the AI indicator mean my work is clean?
It means the detected AI writing fell below the 20% confidence threshold, so Turnitin withheld a figure. That is a low-confidence result rather than a confirmed zero, and it should be read alongside the rest of the report [3].
Can I check my draft without it being stored in Turnitin's database?
That depends on the service. Turnitin0 checks your file without adding it to the student paper database and does not share reports with third-party databases, so a draft check will not later surface as a match against your own final submission.
Will my university accept a pre-submission check?
Many do, but policies differ by institution and by module, so confirm in your handbook or with your tutor before uploading a draft [4]. The check is a revision aid, not a substitute for your own writing.
What file types and sizes can I check?
Turnitin0 accepts English .docx, .pdf, and .txt files between 300 and 30,000 words, under 20 MB, and returns both the AI detection report and the similarity report in one order.