Turnitin AI Writing Indicator

Runs inside an existing workflow with no extra student action, and Turnitin publishes more methodological detail and more explicit caveats than most competitors. The indicator is shown only to instructors, not students.

Not recommended as a decision tool. If your campus has it enabled, the only defensible use is as a private prompt to look more closely at a draft's process evidence — never as the basis for an allegation.

MIT's August 2026 committee report recommends against relying on AI detectors, warning they may mistake the writing of non-native English speakers or neurodivergent students for AI, and that even low false-positive rates put students on edge with serious individual consequences. Turnitin's own under-1% claim applies only to documents already over 20% AI content with 300+ words of prose — it does not describe short or heavily revised writing. Vanderbilt disabled it after calculating a 1% rate would falsely accuse roughly 750 students a year. For an ESL-heavy college the disparate impact lands on the students least able to contest it.