Does a Turnitin AI score mean that percentage chance of cheating?

A percentage looks wonderfully decisive when someone else has to defend themselves against it. Before treating a Turnitin AI score as a verdict, check what the number measures. Turnitin's existing guide describes the percentage as the share of qualifying long-form prose its model identifies as likely AI-generated or AI-generated and then modified. It is not a percentage chance that the student cheated. The AI percentage is also separate from the similarity score. The company explicitly warns that its model can misidentify writing and should not be the sole basis for adverse action against a student. Even correctly identified AI involvement would still need to be considered against the assignment's rules. If a score is being used against you, I'd ask to see the highlighted passages, the rule allegedly broken, and the other evidence supporting that conclusion. Keep any drafts and notes you already have; I wouldn't invent a writing history or rewrite the disputed submission just to make another detector happier. Follow the school's review or appeal process. For a teacher, I'd settle those questions before asking a student to prove innocence by producing a more reassuring number. This is a reading of the vendor's documentation, not a detector test or a finding about any particular student. What should a school have to explain before an AI flag becomes an accusation?