FOI_25-147 Academic misconduct and artificial intelligence
Date of response: 03 June 2025
We have now considered your request of 09 May 2025 for the following information:
I’m writing under the Freedom of Information act - I would be grateful if you are able to provide me with the following information:
Please could I have, for each of the academic years 2019/20, 2020/21, 2021/22, 2022/23, 2023/24 and 2024/25 (to date);
Question 1. The number of confirmed cases of academic misconduct among students.
Question 2. The number of confirmed cases of plagiarism among students.
Question 3. The number of confirmed cases of academic misconduct among students relating to the misuse of Artificial Intelligence / Generative AI / Large Language Model (e.g. Chat GPT or Google’s gemini) tools
To further clarify 3): If you have a flag in your record keeping system which relates to misuse of AI please use that to classify cases. If you don’t have a flag in your record keeping system for misuse of AI please return the number of cases of academic misconduct found against students where one of the following key word phrases was mentioned: “artificial intelligence”, “ ai ”, “chat gpt”, “llm”, “generative ai” or “large language model”. I appreciate AI misuse may not have been classed as academic misconduct for some of the earlier years asked for so please only provide data for academic years where this was the case.
Our response:
We regret that on this occasion it is not possible to provide the requested information.
Under Section 1 of the Freedom of Information Act, we can confirm that the University does hold the information requested, however on this occasion it is not possible for us to provide any of the information relating to the number of confirmed (where the student was subjected to a formal disciplinary investigation into academic misconduct and found to have breached our General Regulations for students) cases of Plagiarism (question 2) in academic years 2019-20 to 2024-25 to the date we received your request, and the number of confirmed cases of academic misconduct which related to misuse of Artificial Intelligence software such as those noted in your request (question 3).
We have determined that the cost of finding and assembling the requested information will exceed the ‘appropriate limit’ as defined by section 12 of the Act and the Freedom of Information and Data Protection (Appropriate Limit and Fees) Regulations 2004/3244.
'The ‘appropriate limit’ of £450, which equates to 18 hours’ work, as defined by the Information Commissioner’s Office, can relate to one request in its entirety or to a series of linked requests. If the University cannot locate, retrieve and extract some or all of the requested information within the 18 hours we are not obliged to retrieve any of the requested information.
Information relating to cases of Plagiarism and those which may have involved the misuse of any Artificial Intelligence, is held within our Learning and Teaching Services (LTS) and our Student Senate Disciplinary Committee (SSDC) departments (where cases were referred for formal disciplinary).
To explain our position, LTS record cases of academic misconduct, and those suspected cases which were ‘confirmed’ following a formal disciplinary investigation (records of such are held by SSDC), as a breach of our General Regulations for Students (Regulation 18- Plagiarism and collusion and Academic integrity). This regulation is not specific to plagiarism alone as it covers plagiarism, collusion and contract cheating) so this would mean a review of the case files to extract those cases solely involving plagiarism.
For the period of your request, there were 1,473 confirmed cases of academic misconduct logged as a breach of Regulation 18 (where the student was found to have breached regulation 18 of the General Regulation for students), which may fall into the scope of your request. The only way of identifying, extracting and recording the exact information you seek would be to interrogate each case file manually.
We have calculated it would take a total of 10 minutes per each ‘confirmed’ student academic misconduct case file to extract the requested information you seek for question 2. We have calculated it would take an estimated 245.5 hours, to locate and manually interrogate each individual academic misconduct case file.
Further to this, to respond to question 3 of your request, all cases of any academic misconduct which involved misuse of AI would require review of all cases of academic misconduct to establish if AI was an attribute.
Information relating to students facing a formal misconduct investigation due to suspected academic misconduct and the nature of those investigations is held within our Learning and Teaching Services division (LTS), however this division does not routinely record the number of academic misconduct cases involving Artificial Intelligence (AI) specifically. There is no sub-categorisation applied to cases of academic misconduct records, which would allow us to clearly separate out those cases involving AI only.
The only way of identifying, extracting, and recording the exact information you seek for question 3, would be to interrogate each individual formal misconduct investigation case file manually, to ascertain which cases involved the suspected use of Artificial Intelligence.
We have identified 453 formal misconduct investigations into just one category of academic misconduct; suspected plagiarism and collusion, for the academic year 2023-24 alone, which may fall into the scope of your request.
We have calculated it would take a total of 10 minutes per each student academic misconduct case file, for those held for 2023-24 year alone, or 75.5 hours, to locate and manually interrogate each individual file. Which exceeds the appropriate time limit as outlined above.
To assist you in formulating a request to which we can respond, we can provide to you the number of confirmed cases of academic misconduct among students, i.e the number of students who were found to have breached Regulation 18 of our General Regulations for students (all cases of Plagiarism, Collusion and Academic Integrity) for each academic year of the requested period.
Please note we have not considered whether any exemptions may apply to such a request.
We should also point out that any revised request you submit will be treated as a new FOI request, and the 20 working-day time-limit will begin again.
We are sorry we cannot provide the data you requested, but trust this response explains our position.