At Dundalk Institute of Technology (DkIT), each assessment brief sets AI use as unrestricted, restricted or prohibited. Where use is permitted you add a declaration of use with your prompts and outputs, and where a submission raises questions you may be invited to an Oral Verification Interview. Here is what DkIT's guide for students and its Academic Integrity Policy say, checked 22 October 2026.
Used AI? If your AI declaration clearly doesn’t cover the use you describe elsewhere in your text, Esmond points it out.
Try the free previewYou get a free preview first. You only pay if you want the full review.
DkIT's guide Generative Artificial Intelligence (AI) and Your Assessments, A Guide for Students (version 1, approved by Academic Council on 29 September 2023) explains that the terms of any permitted use are written into the assessment brief by the lecturer: "The use of artificial intelligence will be “unrestricted”, “restricted”, or “prohibited”." The guide also asks you to check with your lecturer before you use these tools.
It separates using AI to learn from using it for assessed work. It mentions ethical uses such as drawing up a study plan, summarising information, narrowing down a topic or making practice questions, and it adds: "Remember that it is always unethical to use generative AI to produce work that you claim as your own."
A fictional example of a commentYour conclusion says the measure works, but your interviews mainly show how staff experience it. Could you limit the conclusion to what your data can show?
Upload your PDF and the free preview shows you a first comment on part of your text, no account needed. You only pay if you want the full review.
Try the free previewIf you are permitted to use AI and you do, the guide asks for a declaration of use in your submission. It covers five things: a written acknowledgement, the technology you used, the prompts, the outputs, and how you used the output, directly or after changing it. The guide's model declaration also records the dates of the prompts, and long outputs can go in an appendix.
There is a second case that is easy to overlook. If AI is permitted and you choose not to use it, the guide asks you to include this declaration: "No content generated by AI technologies has been used in this assessment". Separately, the coversheet in DkIT's Continuous Assessment Procedures (approved 29 September 2023) includes a statement about generative AI among the points you declare with each continuous assessment.
The Academic Integrity Policy and Procedures (version 6, approved on 28 April 2023) was updated that year to include generative AI tools. Its wording leaves little room: "Using generative artificial intelligence tools (e.g. ChatGPT) in an assessment unless explicitly permitted to do so and with proper acknowledgement, is a form of plagiarism." Both conditions count: permission and acknowledgement.
The policy names its own next review date, 28 April 2026, which has passed. On 22 October 2026 we found no newer version on DkIT's policies page, so ask your programme whether an update is planned.
The guide is frank about how misuse comes to light. Lecturers are subject experts who know your work in class, and they notice inaccuracies, false references and an inconsistent style. On detection software, the guide says the tools are at an early stage and still need development and testing, so a result can serve as one signal among several when a case is looked at more closely.
The guide also points out that what generative AI produces is not always accurate: references can look convincing and still not exist. It makes checking the accuracy of that information your responsibility.
The policy names an oral verification interview as one way a lecturer can verify the authenticity of submitted work, next to Turnitin or another method suited to the discipline. Its guidelines for staff describe how the interview runs: you are told in advance, a lecturer who is alone on the module brings a colleague, and the Head of Department stays away because of a role later in the process. "The student is entitled to bring a supporter to the Oral Verification Interview." The supporter only observes and cannot answer for you.
The questions should stay with your knowledge of the subject of the assessment, and "The student may also be asked to describe how they approached the assessment task and the sources of information they used." After the interview the lecturer does not impose a penalty; if a breach is still suspected, the case usually goes to a Departmental Disciplinary Panel. If you cannot attend, a second interview is arranged.
Esmond evaluates; it does not write. You upload your own draft and get comments in the margin plus a report that shows where your argument, method or use of sources is weaker, and why. It does not write or rewrite your text: it points to the weak spot and offers guidance on how to revise it. Where a comment includes an example, it illustrates a principle; the sentence you submit is one you write yourself.
DkIT has no connection with Esmond and has not endorsed it. Esmond uses language models, so using it counts as AI use. Rules on AI vary by university, programme, module and assignment. Check your module or programme handbook before you upload, and if it is unclear, ask your supervisor or module coordinator and declare the use if your institution asks you to.
This page summarises published institutional guidance and links to the original sources. It is not legal advice and not a ruling on your assignment: what your programme, module or supervisor sets always comes first.
A review costs from €3.99, with a free preview first. The price depends on the length of your document and you see it before you pay. No subscription, no account.
The guide and the policy speak of assessments in general and set no separate AI rule for the dissertation. Check your project brief and ask your supervisor which setting applies.
The guide gives a fixed declaration for that case, quoted on this page, to include in your submitted assessment.
No. DkIT has no connection with Esmond and has not endorsed it. Where an assessment is set to prohibited, do not use Esmond for it. Where AI is permitted, include Esmond in your declaration of use as the guide asks.
Official documents, checked 22 October 2026. Quotations are word for word; the rest is our summary. Read the full text at the source.
Get your text reviewed before you submit
You get a free preview first. You only pay if you want the full review.