Universities That Turned Off Turnitin's AI Detection: the 2026 List

At least 20 universities in five countries have switched off Turnitin's AI writing detector or never turned it on, and every one of them is in the table below with a link to its own notice. The newest are Washington State University, which cancelled the feature in February 2026, and Curtin University in Australia, which turned it off on 1 January 2026. In almost every case the familiar similarity check stays on and only the AI score goes away. One thing worth knowing before you read on: a university switching off the AI score has not said that AI use is allowed. Your course rules still decide that.

The full list

Newest first. The date is when the change took effect, where the university published one. Some universities only say that the feature is off today, so those rows say not dated. The reason column uses the university's own words where it gave any. The last column names the notice, and every link is in the sources at the end of this article.

Why they switched it off

Read the notices side by side and the same four reasons come up again and again. None of them is about cost alone, although Waterloo did mention the price in US dollars. They are about what happens to a student when the score is wrong.

• ✦
  False positives at scaleTurnitin has put its false positive rate at about 1 percent. That sounds small until you multiply it. Vanderbilt sent 75,000 papers through Turnitin in 2022, so 1 percent is about 750 students. Washington State ran the same sum on fall 2024 and got about 1,485 assessments.
• ✦
  Bias against second language writersWaterloo and Northwestern both pointed to research showing that detectors flag writing by people whose first language is not English more often. Plain, careful prose is exactly what a cautious second language writer produces, and it is also what detectors read as machine made.
• ✦
  Their own tests did not hold upWaterloo's technology team tested the tool and found its benefits inconclusive. More than once it scored human writing as 100 percent AI. Berkeley also cites its own test results as the reason it opted out.
• ✦
  Trust between staff and studentsCape Town said continued use risked student trust and fairness. Curtin framed its change around trust and clarity. A score that a student cannot see, cannot check and cannot appeal against with evidence of its own works against both.

What replaced the AI score

Almost nobody replaced one detector with another. The text matching that finds copied sources stays on at Curtin, Washington State and the others, because it does a different job: it finds the exact passage and the exact source, which a person can check. What changed is how universities handle suspected AI use.

Waterloo asks instructors to redesign assessments and teach AI literacy instead. Cape Town folded its decision into an AI in Education Framework that its Senate committee endorsed in June 2025. Curtin points students to Grammarly's authorship feature, which records how a document was written rather than guessing from the finished text. At Washington State, the Faculty Senate took up how instructors should now handle suspected unauthorized AI use. The common thread is evidence a person can look at, not a percentage from a model nobody outside the vendor can inspect.

Who turned it back on

The list only moves in one direction most of the time, but not always. DePaul University turned AI detection off for all Turnitin accounts on 28 September 2023, after its Teaching, Learning and Technology committee asked for more time to evaluate it. Its current guidance says faculty can use Turnitin's AI detector again, framed as a starting point for a conversation rather than a strict enforcement tool. Wright State University is reported by the Programs.com tracker to have switched it off in January 2024 and back on in November 2024. We could not read Wright State's own report, so it is not in the main table.

That is the honest picture. Some universities that switch off the score do come back to it, usually with new rules about how it can be used. When you check your own university, check the date on its page, not just whether it appears on a list like this one.

The UK has moved a step further

British universities mostly made their AI score decision back in 2023, when Turnitin switched the feature on for everyone unless they opted out. Newcastle, for example, says it does not use the AI feature and calls detectors highly unreliable. The 2026 story is bigger. Times Higher Education reported on 9 October 2026 that some universities are leaving Turnitin entirely.

• ✦
  SouthamptonWill end its association with Turnitin when the current deal lapses. The concern was proposed licence changes that would have let student submissions be used to train AI services.
• ✦
  NewcastleWill stop using Turnitin after the 2026 to 2027 academic year, citing the potential use of staff and student data for AI purposes.
• ✦
  WestminsterWill retire Turnitin when its contract ends in July 2027. It said the core plagiarism checker is largely ineffective against the biggest current integrity risk.
• ✦
  Aberystwyth, Aberdeen and CambridgeAberystwyth is under contract until July 2027 and lets uncomfortable students submit through Blackboard instead. Aberdeen is reviewing its use. Cambridge says it will not accept licence changes that affect data privacy or intellectual property.

Turnitin has since postponed the licence changes. Whether the exits go ahead is worth watching, because a university that drops Turnitin drops the AI score with it.

What this means if you are a student

If your university is on the list, an AI percentage should not appear in your case at all. If it is not on the list, the score may still be used, and Turnitin's own guidance says it should never be the only basis for action against a student. Either way, a few habits protect you.

1. 01
   Find your university's noticeSearch your university's site for Turnitin and AI. The table above links the notices we found. A page dated this year beats a list on someone else's site, including this one.
2. 02
   Keep your workingDrafts, version history in Google Docs or Word, notes and sources are the evidence that settles a dispute. The nursing student in the ABC's Australian Catholic University story waited six months to be cleared.
3. 03
   Ask what the evidence isIf you are accused, ask what evidence exists beyond a detector score. ACU told the ABC that any case where Turnitin's AI tool was the only evidence is dismissed immediately.
4. 04
   Know the false positive riskDetectors misread some human writing, most often short, formal or second language text. If you want to see how a detector reads your own work before you submit, a free AI detector shows you the score in seconds.

How we built this list

We started from news coverage, two public trackers and university technology pages, then opened every source ourselves on 23 October 2026. A university made the table only if its own page or a named news outlet says the AI detection feature is off, was switched off, or was never enabled. We quote the reason the university gave and write not stated when it gave none.

We left some names out on purpose. Cambridge and Manchester are often listed as early opt outs, but the evidence is statements to journalists in 2023, not a page we could read. The University of South Australia's notice sits behind a login. Liverpool's guidance at one point described the feature as enabled after an initial pause. Wright State and DePaul are covered in the section on universities that turned it back on. For a longer list of US colleges, the Programs.com tracker counts 38, and we used it to find several of the US notices above.