We've written before about what makes a similarity score "bad", and the AI score deserves the same honest treatment — because it behaves completely differently, and applying similarity-score intuition to it is how students misread their situation. The uncomfortable headline: there is no institutional consensus on an "acceptable" AI percentage, and unlike similarity, there's no innocent-by-default explanation for a high one. But the three results the report can show — 0%, *%, and a stated number — each mean something specific, and knowing how instructors are told to read them is worth a lot.
First: 0% and *% are not the same thing
These two get conflated constantly, and they shouldn't be. A 0% means the detector did not identify any of the document's qualifying text as likely AI-generated. An asterisk (*%) means the detector did register something, but the result landed above 0% and below roughly 20% — a range where Turnitin's own data showed elevated false-positive risk, so it suppresses the exact number rather than state it (Turnitin's report documentation describes the display behavior). Practically, an asterisk is still a good outcome: Turnitin explicitly cautions instructors against acting on that range, and any argument about whether you're "really" 6% or 14% is an argument Turnitin itself refuses to have. But if your report says 0%, you're not in the asterisk zone at all. (More on the report's mechanics in how to read the AI writing report.)
There is no published "safe range" — here's what exists instead
You'll find blogs offering neat bands — "20–40% means this, 70%+ means that." We won't, because no such institutional standard exists to cite: unlike similarity, universities largely don't publish numeric AI-score thresholds, and Turnitin's guidance to instructors is that the indicator alone proves nothing. What can be said honestly is structural:
- A 0% means no qualifying text was identified as likely AI-generated. An asterisk (*%, under ~20) means a low-confidence detection Turnitin declines to quantify — and the range it explicitly tells instructors to treat with extra caution.
- Any stated number means the report is making a confident claim about some share of your prose. From there, what matters isn't where the number sits in an imaginary band — it's which passages are highlighted and whether you can account for every one of them.
- The higher the number, the more of the document that claim covers — and the more likely a marker reads the highlighted text closely and compares it against writing they know is yours. Scrutiny scales with the score; verdicts don't come from it.
If your institution has published actual guidance on AI scores, that document outranks anything here. Most haven't — and that absence is the real answer: at most institutions, no official "acceptable percentage" exists to point to.
Why "acceptable" is the wrong frame
A similarity score has innocent explanations at almost any level — quotes, bibliographies, templates. An AI score offers far fewer, and the main one is the detector being wrong, which absolutely happens (here's the accuracy evidence, and who it happens to most) — but that's a defense you make, not a percentage you point to. So the practical question isn't "what number is acceptable?" It's "can I account for every highlighted passage?" If the flagged text is yours, you can defend it and revise it. If it isn't, no threshold will save you.
If you're checking before submission
This is the score's one great property: it's fully knowable in advance. Run the document through the real Turnitin and you get the same AI report a marker would see — percentage and highlighted passages — with time to rework anything flagged. An asterisk on Tuesday means a calm submission on Friday; a 45% on Tuesday means you found out while it was still an editing problem.