University of Galway said it “does not approve the use of AI detection tools” and assesses suspected academic misconduct “under the procedures described” in its Academic Integrity Policy.
Four students who spoke to SIN asked not to be named because they did not want to be identifiable while discussing details of their academic integrity processes.
The Irish Examiner reported on 23 December 2025, based on Freedom of Information data, that Irish universities had identified “at least 416 suspected cases where students may have used AI to gain an advantage in exams or assessments.”
The same report said UG “identified 298 cases,” the largest share, “where AI was suspected to have been used by students to cheat,” while noting that “gaps remain” because some institutions do not record suspected AI misuse as a standalone figure.
The university was “one of the only universities to collect specific data on cheating with AI,” based on FOI documents, according to the report.
A spokesperson for UG said: “There are no plans at this time for a move toward a standard disclosure requirement, to the best of my knowledge.”
In a statement to SIN, UG Academic Integrity Officer Dr Justin Tonra said the university’s Academic Integrity Policy states that “use of artificial intelligence by a student to generate content which is presented as their own for assessment is academic misconduct.”
Tonra said the Academic Integrity Office has produced “sample statements in assignments and syllabi” to set out “appropriate and inappropriate uses of generative artificial intelligence.”
He added that the office has also produced material on “literacy software and academic integrity.”
Tonra said it has also issued material on “mitigating academic integrity risks in unsupervised written assignments.”
He said other units, including the Centre for Excellence in Learning and Teaching, support staff with assessment design.
“As a proportion of reported academic misconduct cases in the University, the number that involves misuse of GenAI is growing,” Tonra said.
“Suspected cases of academic misconduct are assessed under the procedures described in the Academic Integrity Policy,” he added.
“The university does not approve the use of AI detection tools,” Tonra said.
UG’s academic integrity FAQs state the university “does not authorise or advocate” the use of “AI Detectors” for checking student work.
The FAQ warns staff not to upload student work to third-party detection tools because it “may constitute an improper use of personal data and/or intellectual property.”
UG’s Academic Integrity Policy uses an “accepted standard of proof” and requires the decision maker to decide whether it is “more likely than not” that academic misconduct has taken place.
UG’s Academic Integrity Policy says a “Courageous Conversation” takes place before any formal investigation and involves an Academic Integrity Advisor setting out the allegation and explaining that a “full admission” at that stage means the “most serious outcomes” for that category will not apply.
In a December 2025 policy framework, the Higher Education Authority called on institutions to “prohibit the use of AI detectors and probabilistic tools as determinative evidence of misconduct” and to “ground integrity investigations in natural justice,” including “presumption of innocence” and “timely resolution.”
In practice, that means a detector score should not decide a case on its own, and institutions should rely on fair procedures and evidence when assessing suspected misconduct.
A QQI survey report published in 2025 found 54 per cent of learner respondents were unsure if there are similar rules on using GenAI in assessments across modules, courses and subjects at their provider.
The same report found 46 per cent of staff respondents were “not at all confident” that current GenAI detection tools can reliably detect the use of GenAI by learners in assessed academic work.
One student said she first saw a Canvas notification in late January 2026 when grades were released, although the comment displayed an earlier date.
She said the notification stated her work was under review for suspected generative AI use and raised concerns about citations and the overall style of claims and argumentation.
“It was a bit of a punch in the face,” she said.
The student said she later traced citation problems to a bibliography formatted using a writing tool.
“I use Grammarly to do my bibliography,” she added.
She said she returned to her sources and module materials to show references existed and that the errors were her own.
The student described the issues as “clerical and interpretive errors.” “They could have just said two of your citations aren’t coming up,” she said.
The student said she emailed clarifications and later received an update stating her clarification had been accepted and the essay was treated as a bona fide effort.
A second student said she was told in December 2025 that she had been randomly selected for a brief oral follow up after submitting a mid-term essay for an English module.
“The email explained that I had been randomly selected for a brief oral follow up,” she said.
She said it also stated the meeting “was not related to any breach of academic integrity.”
The student said the lecturer opened the essay during the meeting and asked how she chose her question and texts and how she developed her ideas.
“I wasn’t asked directly about AI, ChatGPT, Grammarly, or any tools,” she said.
She said the lecturer described her work as having “the crooks without the meat” when explaining what needed development.
The student said the lecturer also noted the meetings can discourage misuse because “anyone who wrote their own work should be able to discuss it.”
A third student said she admitted using AI to help frame ideas for an assignment in November 2025 after receiving a message stating it had been detected in her submission.
She said she used it for “basic ideas for what to write,” and said it was not taken “word for word.”
The student said she was offered “resubmission with a cap at 50%” or having marks deducted.
“I told her to dock my marks and that was the end of it then,” she said.
The student said she was informed the outcome “would be written on my permanent record.”
A fourth student said he submitted an assignment in November 2025 a few days before the deadline and only learned it had been flagged when semester results were released in late January 2026.
He said he discovered the issue through Canvas when he checked feedback and found what he described as a “brief” comment stating: “Suspected AI use.”
The student said no further explanation was provided for why the work was suspected.
“To say I was shocked was an understatement,” he said.
“I didn’t sleep well that night,” he added.
The student said he emailed the module coordinator to raise the issue and said the work was reviewed.
He said he received his grade the following day and added: “That doesn’t undo the stress.”
“The current technology flags even the templates we have to use,” he said.
UG Students’ Union vice president for education Seán de Búrca told SIN students are not clear on what AI use is permitted because expectations vary by module and assessment type.
“No, not at all,” he said.
“In some modules it’s encouraged, in others it’s prohibited, and you get everything in between,” he added.
De Búrca said students can assume rules from one module apply elsewhere, which can lead to “mistakes” that end up in integrity processes.
“It’s not so much queries and complaints but moreso people making mistakes and being brought in for academic integrity meetings,” he said.
He said students invited to meetings are told they can bring a support person, including a Students’ Union representative.
“What I like to do is talk to the student first,” he said.
“Honesty is the best policy,” he added.
De Búrca said some students can still be incorrectly flagged under the current approach.
“There will be false positives where there’s no definitive safeguard against,” he said.
He said international students whose first language is not English can be vulnerable where they use AI tools to “brush up the English.”
“The result is that it gets flagged as 100% AI,” he added.
De Búrca said UG could reduce confusion by setting more consistent expectations.
“The most obvious improvement is having a University wide AI policy,” he said.
SIN is preparing a separate report on recent national guidance on generative AI in higher education.
