Plagiarism and AI at Trinity College Dublin: permitted by default, declared on the coversheet

The short answer

At Trinity, generative AI is permitted by default. The Calendar says so in terms: “The use of Generative Artificial Intelligence (GenAI) is permitted unless otherwise stated in programme handbooks.” That is unusual, and it is the opposite of the assumption most students arrive with.

Two conditions ride on it. Where you use GenAI output you must acknowledge and cite it. And a restriction can be imposed at programme or module level — so the real question is not “is AI allowed at Trinity” but “what does my handbook and my module description say”. The coversheet declaration you sign is written to make you answer exactly that.

§97: permitted unless otherwise stated

Trinity’s undergraduate General Regulations live in Calendar Part II, and the Academic Integrity provisions are §§96–100. The Calendar 2026-27 page carrying them was updated on 28 August 2026. Section 97, verbatim and in full:

“97 The use of Generative Artificial Intelligence (GenAI) is permitted unless otherwise stated in programme handbooks. Where the output of GenAI is used in a document or work output, this usage should be acknowledged and appropriately cited, as per Library guidelines on acknowledging and referencing GenAI. Where restrictions are applied around the use of GenAI in particular modules, those restrictions must be explicitly stated in modules descriptions in Blackboard Ultra and/or in programme or module handbooks.”

The section then points onward to the College’s GenAI statement.

Take the three sentences one at a time, because each does separate work for you.

The default is permission. If your handbook and your module description are silent, GenAI use is permitted. You are not obliged to hunt for an authorisation that does not exist.

Acknowledgement is not optional.Permission to use is not permission to be silent. Where GenAI output appears in your work, §97 requires it to be acknowledged and cited, following the Library’s guidance. The Library’s own GenAI guide restates both halves together:

“Please note, the use of Generative AI is permitted unless otherwise stated in course and module handbooks, or in module descriptions in the Blackboard VLE. However, you must (at a minimum) acknowledge its use, using the guidance given in this LibGuide.”

That guide also says something candid about its own status: “GenAI is evolving rapidly and there is not yet general consensus on how to acknowledge and reference it. This guidance will therefore continue to be reviewed and updated.” Check the current version rather than one you read last year. For what a good acknowledgement generally contains, our guide to AI disclosure statements may help — but Trinity’s Library guidance is the form that governs here.

A restriction has to be written down, in a named place. This is the clause that protects you. Restrictions “must be explicitly stated” in module descriptions in Blackboard Ultra and/or in programme or module handbooks. A verbal remark in a lecture, or an assumption that everyone knew, is not what §97 describes. If you are ever told you breached a restriction, the first thing to establish is which handbook or module description stated it, and when.

The Library’s misconduct page lists, among its examples of misconduct, “Content generation using artificial intelligence without acknowledgement”. Note where the offence sits in that sentence. It is the absent acknowledgement, not the use.

The coversheet declaration, in full

This is the most concrete disclosure instrument in this whole series of pages, and it is worth reading before you tick it rather than after. From the Library’s academic-integrity declaration page, last updated 12 June 2026:

“Each coversheet that is attached to submitted work should contain the following completed declaration:

“I have read and understood the academic integrity provisions in the General Regulations of the University Calendar for the current year, found at http://www.tcd.ie/calendar.

I have also read and understood (i) the programme-level permissions regarding the use of GenAI in this assessment task, outlined in the programme handbook that relates to this assessment, and/or (ii) any module-level permissions that relate to this specific module, which are outlined in the module descriptions in the Blackboard VLE (Students should contact their module co-ordinator if they are unsure about the GenAI permissions that relate to this assessment).

Where I have used the output of GenAI in my work, I have acknowledged its use in my work and understood the guidance, located at: https://libguides.tcd.ie/gen-ai.

I have also read and understood the Academic Integrity guide located at https://libguides.tcd.ie/academic-integrity and completed the ‘Ready Steady Write’ Tutorial on avoiding plagiarism, located at https://libguides.tcd.ie/academic-integrity/ready-steady-write.”

Note that Schools can create a local version of the coversheet that stipulates the programme or module-level permissions that pertain to the assessment, if necessary.”

Four separate statements, and each one is a commitment.

• You have read the Calendar’s integrity provisions for the current year — §§96–100 for undergraduates.
• You have read the programme-level and/or module-level GenAI permissions for this specific assessment. The declaration even tells you the remedy if you are unsure: contact your module co-ordinator. Do that before signing, not after.
• You acknowledged any GenAI output you used, following the Library guidance.
• You have read the Academic Integrity guide and completed the ‘Ready Steady Write’ tutorial. This one is a training requirement with a completion step. If you have not done it, do it — you are declaring that you have.

The last line matters practically: your School may issue its own version of the coversheet stating the permissions that apply to that assessment. Where a local coversheet exists, it is the most specific statement of the rule you will get, and it is worth keeping a copy of the one you actually submitted under.

What counts as academic misconduct

Calendar Part II §96 sets the general test, verbatim:

“96 General
It is clearly understood that all members of the academic community use and build on the work and ideas of others. However, it is essential that we do so with integrity, in an open and explicit manner, and with due acknowledgement.
Any action or attempted action that undermines academic integrity and may result in an unfair academic advantage or disadvantage for any member of the academic community or wider society may be considered as academic misconduct. Examples of academic misconduct can be found in the curriculum glossary.”

It is a principle rather than a list, and it reaches attempts as well as acts. The Library’s working definition of plagiarism is narrower and more familiar: “Presenting work/ideas taken from other sources without proper acknowledgement. Submitting work as your own for assessment or examination, which has, in fact, been done in whole or in part by someone else”.

Group work has its own section, §98:

“Students should normally submit assessments and/or examinations done in co-operation with other students only when the co-operation is done with the full knowledge and permission of the lecturer concerned. … When work is submitted as the result of a group project, it is the responsibility of all students in the group to ensure, so far as is possible, that no work submitted by the group is plagiarised, or that any other academic misconduct has taken place. … Should a module co-ordinator suspect academic misconduct in a group assignment, the procedure in cases of suspected academic misconduct must be followed for each student.”

Two consequences. The responsibility for a group submission is shared, so a co-author’s undeclared AI use is your problem too — agree the GenAI rules inside the group early, in writing. And if misconduct is suspected in group work, each student goes through the procedure individually; you are not judged collectively.

Section 99 places a duty on your School as well as on you:

“99 Avoiding academic misconduct — Students should ensure the integrity of their work by seeking advice from their module co-ordinator, tutor or supervisor on avoiding academic misconduct. All schools and disciplines must include, in their handbooks or other literature given to students, guidelines on the appropriate methodology for the kind of work that students will be expected to undertake…”

If your handbook contains no such guidelines, that is a gap on the School’s side of §99, and worth raising rather than working around.

Academic misconduct also appears in the disciplinary offences list at Part II §III(1), “CONDUCT AND COLLEGE REGULATIONS”: “(g) academic or research misconduct, including, but not limited to plagiarism”.

Four levels, and what each one records

Trinity publishes an “Academic Misconduct Levels and Consequences” table, with an Academic Integrity Form (Calculator) producing an indicative score. The four levels:

• Level 1 — Poor Academic Practice/Conduct. Indicative score 1–200. “Mandatory academic integrity training is required.” Recorded on SITS, not on the transcript. You may accept a referencing penalty or correct the work, which is then reassessed.
• Level 2 — Academic Misconduct (Minor). Indicative score 201–350. Mandatory training; recorded on SITS, not on the transcript; work must be resubmitted with corrections, and “The component/assessment/examination mark will be reduced or capped at the pass mark.”
• Level 3 — Academic Misconduct (Major). Indicative score 351–500. Mandatory training; SITS record, not on the transcript; work resubmitted at reassessment, and “The mark for the module will be capped at the pass mark.”
• Level 4 — Academic Misconduct (Severe). Indicative score 501+. “Case referred directly to the Junior Dean” for disciplinary proceedings.

The transcript point is worth stating plainly, because students worry about it more than about anything else: Levels 1 to 3 are recorded in SITS and explicitly not on the transcript. A first, low-level finding — poor referencing, an unacknowledged GenAI use caught early — carries training and an academic penalty, and does not follow you onto the document you show an employer or a doctoral admissions committee.

That is genuinely reassuring, and it is also the reason the internal record exists: Trinity keeps the history so that a second instance is treated as a second instance. Its staff guides say so. A “Recording decisions in SITS” guide notes “this applies to all levels, including level 1”, and an “Academic Misconduct Lookup” report checks “if a student has had a previous recorded case of academic misconduct recorded in SITS” and “how many instances of recorded academic misconduct and the corresponding level”.

Level 4 leaves the table entirely. Exclusion and expulsion are not in it; they sit with the Junior Dean under the Consolidated Statutes. Whether a degree can be revoked is not something Trinity documents in the sources we read, and we are not going to guess.

The procedure and the Junior Dean

Calendar Part II §100, verbatim:

“100 Procedure in cases of suspected academic misconduct
If academic misconduct as referred to in §96 above is suspected, the procedure in cases of suspected academic misconduct, available at www.tcd.ie/academic-affairs/academic-integrity must be followed.
The case may be referred to the Junior Dean if:
(i) the facts of the case are in dispute;
(ii) the regulations of the College were not properly applied in the student’s case;
(iii) the Director of Teaching and Learning (Undergraduate), or designate, forms the view that that the consequences provided for under the academic misconduct procedure are inappropriate given the circumstances of the case.
The Junior Dean will interview the student and may implement the procedures as referred to under CONDUCT AND COLLEGE REGULATIONS §2. Nothing provided for under the procedure in cases of suspected academic misconduct diminishes or prejudices the disciplinary powers of the Junior Dean under the 2010 Consolidated Statutes.”

The doubled “that that” is in the original. More importantly, note ground (i).

If you dispute the facts, the case can leave the routine academic route by regulation and go to the Junior Dean, who will interview you. That is a real fork, and it has a trade-off you should understand before you take it. Contesting an allegation moves you into a disciplinary process with an interview and the Junior Dean’s statutory powers, rather than the levels table. It is the right route if you did not do what is alleged. It is not a light procedural step.

This is exactly the situation where your own record of working carries the weight — drafts, dated version history, notes, search history, the earlier messier paragraphs. See how to prove you wrote it yourself and what to do when you are accused. If English is not your first language, it is also worth reading how detectors treat non-native writing, because that is a point you may have to make yourself.

If a detector score is what started this, our guide to an AI detector flagging your thesis sets out what that number does and does not establish on its own.

The decision-makers named on Academic Affairs’ own flowchart are the Dean of Graduate Studies (DGS), the Director of Teaching and Learning (DTL, undergraduate or postgraduate), the Module Coordinator (MC) and the Senior Lecturer (SL).

Postgraduate students: the equivalent Calendar provisions are in Part III, §§1.50–55, per the Library guide’s own citation. We quote Part II here because that is what we read directly; check Part III for your own regulations.

Standard of proof: Trinity does not state one in the documents we retrieved — not in the Calendar, the Academic Integrity Policy (v1.2, 29 May 2026) or the misconduct Procedure (v1.0, 2 September 2024). We could not establish it, and we will not supply a number that Trinity has not published.

The College Statement on GenAI

Trinity is the only institution in this series with a dated, versioned, institution-wide GenAI statement: the College Statement on Artificial Intelligence and Generative AI in Teaching, Learning, Assessment & Research, Version 2, dated 4 June 2026, published through Academic Affairs. It describes itself as provisional:

“This statement is a living document that will be regularly reviewed and updated as Artificial Intelligence (AI) and Generative Artificial Intelligence (GenAI) technologies evolve, and as other related College policies are published.”

Its posture is two-sided rather than defensive:

“As Ireland’s leading university and as a world leader in AI research, Trinity recognises that AI and GenAI offer new opportunities for teaching, learning, assessment and research. We also recognise that these technologies present significant and increasing challenges and risks, including to academic integrity, ethics, privacy, impartiality, intellectual property and sustainability.”

And it explains why the rule lives in your handbook rather than centrally:

“Trinity’s goal is to enable overall consistency in the regulation of GenAI usage, while also respecting where disciplines or degree programmes require specific restrictions in GenAI usage in assessment preparation and execution. … Such regulation could range from how student GenAI usage is acknowledged or cited within student assessment submissions, to prohibition of GenAI usage in the production of student assessment submissions.”

Its two named principles for students are Transparency and Honesty — “We will strive to be transparent and honest in our use of AI and GenAI, and to give clear, fair, impartial and appropriate acknowledgement and citation to any role that AI and GenAI have played in scholarship, teaching, learning, assessment and research” — and Responsibility and Accountability, which stresses “the importance and necessity of human oversight”.

The guidance-for-students section, verbatim in full:

“Students need clear guidance on how to use GenAI in their learning, assessment and research. These guidelines are essential to maintain academic integrity, human oversight, and critical thinking in any output. Students should be guided to an ethical, responsible, and diligent human-AI collaboration. In any AI and GenAI uses for their academic work, students must acknowledge and reference their uses following the guidelines below.”

The statement takes its definition of AI from the EU AI Act (2024) and cites as its context the principles of the Higher Education Authority, the Irish Government’s Guidelines on Artificial Intelligence for the Public Sector, and the OECD Digital Education Outlook 2026.

One finding about this document is worth recording plainly, because of what it tells you about where Trinity has put its attention. The statement does not mention detection at all. We searched its full text for “detect”, “Turnitin”, “similarity”, “Ouriginal”, “false positive” and “accus”: zero occurrences. The same is true of the Academic Integrity Policy and the misconduct Procedure. The whole regime is built on acknowledgement, AI literacy and human oversight.

Do not over-read that. It is not a statement that detection is not used, and it is not a promise that no module runs a checker. It means Trinity has not published a position, so you cannot rely on one either way. For our part: our published error rates are measured on English and German corpora, and no detector — ours included — produces proof. See how accurate AI detectors really are.

On tooling and data protection, the one concrete statement in the document concerns a College-provided service:

“All Trinity staff and students can access Microsoft Copilot Chat through their institutional Microsoft 365 account as part of Trinity’s Microsoft suite. When users sign in with their Trinity credentials, Copilot Chat applies Microsoft’s enterprise data protection controls, indicated by a green shield on screen. This means that all prompts and responses stay within the Microsoft 365 service boundary…”

That is a useful thing to know if you are choosing which tool to use for permitted work: the institutional route has a stated protection boundary, and a consumer chatbot does not. It says nothing about detection or about submitted assessments.

What Trinity does not publish

• Any named text-matching or similarity product.None is named in the Calendar, the Academic Integrity Policy, the misconduct Procedure, the GenAI Statement or the Library’s Academic Integrity guide. The Trinity IT Services URL we probed for Turnitin returned 404.
• Whether an AI-writing detector is enabled anywhere. No position published; see the section above.
• Any opt-out from, or alternative submission route around, detection. Not documented.
• Any statement about where submitted student work is stored or processed. Trinity is in Ireland and therefore inside the GDPR, but no data-residency statement for submitted assessments was published in the sources we read. The Copilot enterprise-boundary paragraph is about a College-provided chat tool, not about assessment. We are not going to infer a guarantee from the jurisdiction.
• Any student self-check. There is no similarity self-check area we could find. What Trinity provides student-facing is the ‘Ready Steady Write’ plagiarism tutorial, which the coversheet declaration requires you to complete — training, not a check on your own text.
• A standard of proof, and aggregate annual case counts. Every level including Level 1 is recorded in SITS, but no totals are published.

One clause in the Calendar does bear on material being fed toAI systems, in the library and copyright context: “…by other means, including digital copying and uploading into any artificial intelligence application.” It is a reminder that what you upload — including other people’s copyrighted material — is itself governed.

Before you submit

• Read your programme handbook and your module description in Blackboard Ultra. Under §97 that is where any restriction must be explicitly stated, and it is what the declaration says you have read.
• If you are unsure, ask your module co-ordinator — the declaration itself names that as the step, so take it before you tick the box.
• Acknowledge any GenAI output, using the Library guidance. Permission to use is not permission to stay silent, and the misconduct example is “without acknowledgement”.
• Complete the ‘Ready Steady Write’ tutorial. You are declaring that you have.
• Agree the GenAI rules inside any group, in writing. Under §98 the responsibility for the submission is shared.
• Keep your drafts and version history. If the facts are ever disputed, §100(i) sends the case to the Junior Dean, and your working record is what you bring.

If you want a read on your own text before you hand it in, our AI check is free up to 1,500 words, with no account and no name.

Sources

• Trinity College Dublin: Calendar Part II — General Regulations and Information — Academic Integrity §§96–100, Calendar 2026-27; page updated 28 August 2026. The Calendar is College’s own regulation. Postgraduate equivalent: Calendar Part III §§1.50–55.
• Trinity College Dublin: Academic Integrity Policy — Policy No QPOLAcI, Version 1.2, Date of Issue 29 May 2026; cites the Statement on Academic Integrity as approved by Council in 2022.
• Trinity College Dublin: Procedure in cases of suspected academic misconduct — Procedure No QPOLAcI, Version 1.0, Date of Issue 2 September 2024.
• Trinity College Dublin: College Statement on Artificial Intelligence and Generative AI in Teaching, Learning, Assessment & Research — Version 2, Date of Issue 4 June 2026, published via Academic Affairs.
• Trinity College Dublin: Academic integrity — Academic Affairs; parent page updated 10 July 2026. Includes the Academic Misconduct Levels and Consequences table and the Academic Integrity Form (Calculator).
• Library of Trinity College Dublin: Academic integrity — declaration — the coversheet declaration; last updated 12 June 2026.
• Library of Trinity College Dublin: Generative AI guide — acknowledging and referencing GenAI; last updated 10 June 2026.
• Library of Trinity College Dublin: ‘Ready Steady Write’ tutorial — the plagiarism tutorial the declaration requires.

This page is orientation, not legal advice. What binds you is your programme handbook and module description, the General Regulations in the University Calendar (Part II §§96–100 for undergraduates, Part III §§1.50–55 for postgraduates), the Academic Integrity Policy and the Procedure in cases of suspected academic misconduct.