In August 2025, Turnitin added a feature built specifically to catch text that's been run through an AI "humanizer" tool — like Undetectable.ai — to evade AI detection. Here's what Turnitin says the feature does, what Undetectable.ai claims right back, and what's actually been independently verified on either side (short answer: not much).
Trusted by students at
In August 2025, Turnitin rolled out a feature it calls AI bypasser detection, built into its existing AI writing detection rather than sold as a separate product. It's designed to flag text showing the fingerprint of having been run through a humanizer or paraphrasing tool specifically to make AI-generated writing read as more human — a step beyond just flagging plain AI-generated text on its own.
Turnitin's own announcement doesn't publish an accuracy or detection-rate figure for the feature. The company's public framing leans more on the arms race than a specific number: Chief Product Officer Annie Chechitelli said plainly that "these companies exist to profit from students' misuse of AI by providing free and easy access to humanizers to conceal AI-generated content."
Undetectable.ai, one of the best-known AI humanizer tools, markets itself directly against this — its own blog states the tool "maintains over a 95% success rate against major platforms like Turnitin and GPTZero." That figure comes from Undetectable.ai itself, not from Turnitin, GPTZero, or any outside test, so it's worth reading as a marketing claim from a company selling the product it's describing.
Neither company's number has been independently audited or tested by a neutral third party as far as we could find. Any "X% success rate" claim for a humanizer tool, this one included, is worth treating as a vendor talking about its own product until proven otherwise.
Turnitin's own materials are honest that AI detection generally isn't perfect: its FAQ says it "advocates for giving students the benefit of the doubt" when a false positive comes up, though it doesn't publish a specific false-positive rate for the bypasser detection feature itself.
We couldn't find a rigorous, independent test — from a university, a research group, or anyone without a stake in either company's outcome — that actually ran humanized text through Turnitin's bypasser detection and published real results. Until one exists, the honest answer to "does Turnitin detect Undetectable AI" is: Turnitin says its new feature is built to, Undetectable.ai says it usually gets past detectors including Turnitin, and nobody outside either company has published a number either way.
Check your writing across multiple detection methods — free, no account needed for your first check.
Check My Essay Free →