Institution-licensed integrity and assessment suite whose AI writing detection is an Originality add-on, reporting a document AI percentage to instructors only.
An institution that already runs similarity checking and needs an AI writing signal inside the same report and the same LMS workflow, with administrator-level controls over whether it runs at all. Turnitin is unusual in publishing its own numbers and limits: a document false positive rate, a sentence false positive rate, file requirements, suppressed low scores, a named list of LLMs the English model is trained to detect, and a dated inventory of which AI provider powers each product feature. It also publishes an explicit FERPA school-official position and, for the EU, states it may be possible to store submissions exclusively in Germany.
Do not buy it expecting a verdict. Turnitin itself states it does not make a determination of misconduct and that the AI writing percentage should not be used as the sole basis for action or a definitive grading measure. It is also wrong for anyone needing a budget figure without sales contact, since no pricing appears anywhere on turnitin.com and AI writing detection is an Originality add-on to Feedback Studio or Similarity rather than an included feature. Institutions that want zero model training on their submissions must leave the feature off: the Trust Center states that where Originality with AI detection is enabled, current student data is anonymised and may be used to train Turnitin's classifier model.
Pros
Cons
These are Turnitin's own published claims, not independently audited results. Turnitin states a document-level false positive rate under 1% for documents with over 20% AI writing, and a sentence-level false positive rate of around 4%. It says it tests over 700,000 pre-ChatGPT academic papers before each model release to validate the document figure, and accepts missing some AI text to hold that rate.
Turnitin says no. Its FAQ states Turnitin does not make a determination of misconduct and that the percentage should not be used as the sole basis for action or a definitive grading measure. Its blog adds that the false positive rate is not zero, so instructors must apply professional judgement, and advises using a highlighted sentence to start a conversation rather than draw a conclusion.
It depends on the licence. The Trust Center states Turnitin trains no generative models or LLMs, only a classifier for AI detection in the Originality licence. It states that if you do not have Originality with AI detection enabled, your current student data is not used for training, and that if you do, your current student data is anonymised and may be used to train the classifier.
The services privacy policy states Customers subject to FERPA contract with Turnitin as a School Official with a legitimate educational interest under FERPA §§ 99.31(a)(1). COPPA is not named; the Children's Privacy section states that where a Customer uses the services with students under 13, or under the local age of consent, Turnitin relies on the Customer to obtain verifiable parental or guardian consent. No minimum age is set by Turnitin.
Facts sourced from the vendor’s own site, last verified 2026-10-10.
The Weekly Source
One email a week. New listings, repriced tools, and the ones we would skip - including this one if its pricing changes.
