Which UK Universities Use Turnitin’s AI Detector?

Most UK universities that have taken a public position have chosen not to run Turnitin's AI writing indicator — but the picture is more varied, more provisional, and worse-documented than the "banned" lists circulating online suggest. This tracker records what each university has actually said, on its own website, with a link and a date for every entry. New to the indicator itself? Start with does Turnitin detect AI? — how the report works before you look up your university's stance.

As of 28 August 2026, of 95 UK institutions reviewed: 22 have an AI-writing indicator explicitly switched off — twenty-one have opted out of, disabled, or do not routinely use Turnitin's indicator (Strathclyde goes further and prohibits submitting work to it), and one (Bath) has the equivalent feature disabled in its Inspera Originality system; 15 more state that AI-detection tools generally are not used, not permitted, not relied on or not recommended (without naming a specific indicator); 7 confirm an AI writing report is in use — Bristol, Northumbria, Leeds Beckett, Birkbeck, Abertay and Birmingham City routinely, and UEA restricted to plagiarism officers; and 51 have no public statement on their own domain (7 reviewed in detail in August, 44 from a wider sweep on 23 August).

This page is maintained quarterly. It reports only primary sources: statements on each university's own domain. If you find an error or a newer statement, contact details are at the end — corrections are credited.

How to read this tracker

Universities express their positions in two distinct ways, and collapsing them overstates the evidence — so this tracker doesn't.

Indicator-specific statements name Turnitin's AI writing detection feature and say whether it is enabled. General detection statements say the institution does not use, permit, or recommend AI-detection tools, without naming the Turnitin feature. A general statement usually implies the indicator is off, but "implies" is not "states", so the two are labelled differently below.

Stance labels: Opted out — the university explicitly states it opted out of or did not enable the indicator. Disabled — it was enabled or piloted, then switched off. Enabled — the AI writing report is in live use. Detection not used/allowed (general) — an institutional statement that AI-detection tools are not used or permitted, without naming the indicator. Every entry carries the date shown on its source, because a 2023 statement is weaker evidence of today's position than a 2026 one.

One more thing this tracker is not: it is not a list of universities that "banned AI detectors". Several third-party lists use that framing; checked against the universities' own pages, the institutions' own word is almost always "opt out" — often explicitly provisional.

The tracker

A note on Cambridge and Manchester: both were widely reported in 2023 to have opted out of the Turnitin indicator specifically. Those reports trace to statements made to journalists, not to pages on the universities' own domains, so this tracker records only what their current published guidance says. If either university publishes an indicator-specific statement, this entry will be upgraded.

Check a specific university

Each university below has its own page with the full statement, the source link and its date, and what the position means in practice for students and staff there.

• Does University of Bristol use AI detection? Yes
• Does King's College London use AI detection? No
• Does University of Greenwich use AI detection? No
• Does University of Nottingham use AI detection? No
• Does University of Glasgow use AI detection? No
• Does University of Birmingham use AI detection? No
• Does Durham University use AI detection? No
• Does University of Liverpool use AI detection? No
• Does University of Warwick use AI detection? No
• Does Newcastle University use AI detection? Not Routinely
• Does University of Bath use AI detection? No
• Does Bournemouth University use AI detection? No
• Does University of South Wales use AI detection? No
• Does University of Manchester use AI detection? No
• Does University of Oxford use AI detection? No
• Does University of Sheffield use AI detection? No
• Does UCL use AI detection? No
• Does Queen's University Belfast use AI detection? No
• Does University of Cambridge use AI detection? No
• Does Northumbria University use AI detection? Yes
• Does Leeds Beckett University use AI detection? Yes
• Does Birkbeck, University of London use AI detection? Yes
• Does Abertay University use AI detection? Yes
• Does Birmingham City University use AI detection? Yes
• Does University of East Anglia use AI detection? Yes
• Does University of St Andrews use AI detection? No
• Does University of Strathclyde use AI detection? No
• Does University of Reading use AI detection? No
• Does University of Edinburgh use AI detection? No
• Does Manchester Metropolitan University use AI detection? No
• Does Goldsmiths, University of London use AI detection? No
• Does Aberystwyth University use AI detection? No
• Does Brunel University of London use AI detection? No
• Does Ulster University use AI detection? No
• Does Imperial College London use AI detection? Not Routinely
• Does University of Leeds use AI detection? No
• Does Edinburgh Napier University use AI detection? No
• Does University of the Arts London use AI detection? No
• Does Middlesex University use AI detection? No
• Does University of Sussex use AI detection? No
• Does City St George's, University of London use AI detection? No

Universities with no public statement

For completeness — and because absence of a statement is itself useful information — the following institutions were reviewed and no public position on AI detection was found on their own domains as of 23 August 2026: University of Edinburgh (Bayes Centre staff guidance stating the university had "decided to opt out" of Turnitin's AI detection has since been removed from ed.ac.uk; with no live statement, Edinburgh is listed here until a current one appears), Cardiff University (Turnitin pages pre-date the AI indicator), University of Exeter (disclosure-based approach; AI FAQs silent on detection), University of Southampton, University of York (describes Turnitin as text-matching only), LSE (its Turnitin policy treats the tool purely as text-matching; Eden Centre guidance cautions against relying solely on detection software, but no indicator status is stated), and Queen Mary University of London (a 2023 announcement title suggests the preview was made available, but the page is login-gated and unverifiable).

A wider sweep on 23 August 2026 of 44 further institutions found no statement on AI detection on their own domains: Aberdeen, Anglia Ruskin, Aston, Bangor, Bradford, Chester, Coventry, Cranfield, De Montfort, Derby, Essex, Glasgow Caledonian, Heriot-Watt, Hertfordshire, Huddersfield, Hull, Keele, Kent, Kingston, Leicester, Liverpool John Moores, Loughborough, Nottingham Trent, the Open University, Oxford Brookes, Plymouth, Queen Margaret, Robert Gordon, Roehampton, Royal Holloway, Salford, SOAS, Staffordshire, Stirling, Sunderland, Surrey, Swansea, Teesside, UCLan, the University of the West of Scotland, Westminster, Wolverhampton, Wrexham and York St John. Several reserve a general right to use detection software or warn that detection "may be utilised" without confirming any tool is enabled (Lancaster, Cardiff Metropolitan, Sheffield Hallam); those are not counted as positions.

No public statement does not mean no detection — some of these institutions may run the indicator without saying so publicly. It means exactly what it says: nothing on their own domain takes a position.

Why so many universities have said no

The stated reasons are strikingly consistent across the primary sources above: unverifiable outputs from closed systems (Durham, Warwick), false-positive risk (Greenwich, KCL, Birmingham, Sheffield, QUB), demographic bias against non-native English writers (Durham, Liverpool), and reliability findings from internal testing (Nottingham, Liverpool).

Those institutional judgements track the independent research. Weber-Wulff and colleagues' 2023 study in the International Journal for Educational Integrity tested 14 detection tools across 756 document tests and found none exceeded 80% accuracy, concluding the tools cannot be recommended as the sole basis for a misconduct case. Liang and colleagues' Stanford study in Patterns (2023) found seven detectors misclassified over half of a set of essays by non-native English speakers as AI-generated. In England and Wales, the Office of the Independent Adjudicator's 2024 guidance points the same direction: allegations must be evidenced to the balance-of-probabilities standard, and Newcastle's academic-integrity guidance explicitly argues an unsubstantiated AI score cannot meet it.

None of this means AI writing detection is useless. It means the sector has largely concluded that a percentage score, on its own, is not evidence — which is also the position this site takes about its own tool.

What this means for students and educators

If you're a student: whether your university runs the indicator is a matter of institutional policy, and as the table shows, at most institutions with a public position, it is off. If your work is questioned, the process that matters is your university's academic-misconduct procedure — our guide to UK academic-misconduct procedures and AI evidence explains how those investigations actually work and what rights you have.

If you're an educator: the sector consensus is that detector output is a signal for human review, never proof. If you use any detector — including ours — the responsible pattern is the one Bristol's guidance describes and our accuracy methodology documents: a starting point for a conversation, weighed alongside drafts, version history, and your knowledge of the student's work. Our guide to AI-detection false positives covers the failure modes to keep in mind.

Methodology and corrections

Every stance in this tracker is sourced to a page on the university's own domain, fetched and read in full, with the date shown on the page recorded. Statements made to journalists, third-party "banned" lists, and vendor claims are excluded — several widely-circulated third-party classifications did not survive checking against the universities' own pages. Entries distinguish indicator-specific statements from general detection statements, and old statements are flagged rather than silently presented as current.

The tracker is reviewed quarterly (next review: November 2026). If you work at a UK university and your institution's position has changed, or you've found a public statement we've missed, email support@genutext.ai with the URL — corrections and additions are credited if you'd like them to be.

You are welcome to reuse this table with attribution and a link to this page (CC BY 4.0).

Turnitin questions students ask

The tracker tells you whether your university runs Turnitin’s AI indicator. These guides answer the questions that usually come next:

• Does Turnitin detect AI? How the report works — the 300-word floor, the asterisk band and the institutional toggle, with sources.
• Can Turnitin detect ChatGPT? — what the AI writing report can and cannot show.
• How accurate is Turnitin’s AI detection? — vendor claims beside independent evidence, every figure dated.
• Can Turnitin detect paraphrased or rewritten AI writing? — how paraphrasing changes the signal, with sourced evidence.
• What an AI-detection percentage means (is 30% bad?) — how to read any AI score before you act on it.
• Check your essay for AI before you submit — an honest pre-submission check that is not Turnitin, and says so.

Frequently asked questions

Does Turnitin's AI detector run automatically at every UK university?

No. The AI writing indicator is a feature institutions choose to enable, and most UK universities with a public position have chosen not to — the table above links each university's own statement. Seven institutions in this tracker confirm it is in use: Bristol, Northumbria, Leeds Beckett, Birkbeck, Abertay and Birmingham City routinely, and UEA for plagiarism officers only.

Can a university accuse a student of AI use based on a detection score alone?

Published guidance points strongly against it. Universities that have taken a position consistently state that detector output is unreliable as sole evidence, and the OIA's good-practice framework requires allegations to be properly evidenced. In practice, institutions rely on broader indicators — drafts, version history, the student's ability to discuss their work — rather than a percentage.

My university isn't listed, or is listed as "no public statement". How do I find out its position?

Check your university's academic-integrity, library, or teaching-support pages for generative-AI guidance — that's where every statement in this tracker was found. If you can't find one, your students' union advice service or course handbook may say. And if you do find a public statement, send it to us and we'll add it.

Is "opted out" the same as "banned AI detectors"?

No, and the difference matters. Most universities describe a provisional decision not to enable one feature, often explicitly under review — Greenwich commits to keeping its decision under review, Nottingham reaffirmed its position only after internal testing, and Bristol shows positions can differ entirely. Several institutions have separately barred staff from uploading student work to third-party detectors, which is a data-protection decision as much as a reliability one.

Which AI detector do most UK universities use?

Turnitin is the detection platform most UK universities license, but its AI writing indicator is a separate institutional switch: in this tracker, more universities with a public position have the indicator off than on, and several say AI-detection tools generally are not used or relied on. Similarity (plagiarism) checking remains in routine use almost everywhere.

Sources and further reading

• University of Bristol: Digital Education Office update · Nov update
• King's College London: KCL AI guidance
• University of Greenwich: University article
• University of Nottingham: 2023 statement · 2024 update
• University of Glasgow: AI software update
• University of Birmingham: GAI guidance
• Durham University: DCAD FAQs
• University of Liverpool: University news
• University of Warwick: ARO news
• University of Dundee: AI in teaching and assessment · Student LibGuide
• Newcastle University: Library FAQ · Integrity blog
• Bath Spa University: AI in Teaching · Digital Fluency guide
• University of Bath: Introduction to Inspera Originality
• Bournemouth University: FLIE announcement
• University of South Wales: USW statement (PDF), linked from the CELT AI Hub
• University of Manchester: AI Hub guidelines
• University of Portsmouth: CADI staff guidance
• University of Oxford: The question of AI detection
• University of Sheffield: AI and academic integrity
• UCL: GenAI and academic integrity
• Queen's University Belfast: QUB guidance (PDF)
• University of Cambridge: Blended learning FAQ · Assessment guidance
• Northumbria University: Turnitin AI writing detector (PDF)
• Leeds Beckett University: Principles for the use of Generative AI (PDF) · Generative AI and academic practice
• Birkbeck, University of London: AI guidance for students 2025–26 (PDF)
• Abertay University: Turnitin AI Detection Tool
• Birmingham City University: AI Tools: BCU Guidelines
• University of East Anglia: Generative AI Policy for Teaching and Learning (PDF)
• University of St Andrews: Generative AI FAQs for students (PDF) · Good academic practice
• University of Strathclyde: Guidance on using Turnitin (PDF)
• University of Reading: Academic integrity and generative AI tools (PDF) · GenAI for teaching and learning
• University of Edinburgh: SADIE project report (Information Services)
• Manchester Metropolitan University: Embracing generative AI responsibly
• Goldsmiths, University of London: Using GenAI responsibly (LibGuide)
• Aberystwyth University: Important update on generative AI for staff
• Brunel University of London: WISEflow: Similarity and related information
• Ulster University: Guidance on use of AI for staff
• Imperial College London: Frequently requested information
• University of Leeds: AI for educators
• Edinburgh Napier University: Position on Generative AI (PDF)
• University of the Arts London: AI and Teaching
• Middlesex University: Generative AI in Learning, Teaching and Assessment policy (PDF)
• University of Sussex: AI and academic integrity
• City St George's, University of London: Use of Generative AI Tools in Doctoral Research
• Weber-Wulff, D. et al. (2023). Testing of detection tools for AI-generated text. International Journal for Educational Integrity, 19:26.
• Liang, W. et al. (2023). GPT detectors are biased against non-native English writers. Patterns, 4(7).
• Office of the Independent Adjudicator (2024). Good Practice Framework.
• Genutext: Accuracy methodology · UK academic-misconduct procedures and AI evidence · AI-detection false positives explained

Turnitin is a trademark of Turnitin, LLC. Ouriginal is a trademark of its respective owner. Genutext is not affiliated with, endorsed by, or connected to Turnitin, Ouriginal, or any university listed. All statements are quoted or summarised from the linked public sources as of the review date; institutional positions can change.

Last reviewed: 28 August 2026