A declaration on the use of artificial intelligence precedes a student’s submission, assuming a cloud commonplace within students’ processes; and certainly, with emerging tools that do not wait for a writers’ honesty, this nimbus of promised efficiency forms a power struggle over higher education’s head. The question now is whether or not AI clouds our own educational judgement and if artificial intelligence itself is the new ideal of what a student should and could be. Trinity’s IT Services lists Turnitin among the tools running inside Blackboard modules. The moment a student submits an essay on Blackboard, a machine reads it before a lecturer does. When a lecturer sets up an assignment through Turnitin, each essay is directed straight to the software, putting it up against databases of webpages, journals, and even past student papers, eventually producing a similarity report with an AI score. By the time a lecturer has opened the student’s submission for the first time, they not only see the document, but are given a rundown of everything wrong with it. AI promises that we can be shown and taught anything, at any time, with no one there at all. If artificial intelligence can then teach us everything, what can we still learn from each other?
Much of it embeds itself without our asking. Trinity’s own College Statement on GenAI admits that, because these tools are now built into everyday devices and platforms, banning them is impossible. Every Trinity student can already use Microsoft Copilot Chat through their college account. The Higher Education Policy Institute (HEPI)’s 2026 survey of 1,054 UK undergraduates found that 95 per cent use AI in some form and 94 per cent use it for assessed work. The share of students pasting AI-generated text straight into their work rose from 3 per cent in 2024 to 12 per cent in 2026. In Ireland, 44.9 per cent of people had used generative AI in the previous three months, according to Eurostat, against an EU average of 32.7 per cent. Almost half of the students HEPI surveyed say AI has improved their experience, mostly by saving time and explaining things quickly. Yet the same survey shows a student body split down the middle. 37 per cent say their university encourages AI, and 36 per cent say it does not. One in five say AI makes them lonelier, while about the same number say it makes them less lonely. In practice, much of artificial intelligence responsibility falls on the student, who is expected to fact-check whatever the tools produce, to seek out primary sources, and to cite every use with the date, the tool and the prompts involved, placing any copied text in quotation marks. Trinity leaves each School free to set its own rules, which may range from simple citation to an outright ban, so two students sitting in the same lecture hall can be held to quite different standards depending on what their course handbook says.
Curiously, the document never mentions Turnitin or AI detection, even though essays submitted through a Turnitin assignment are processed by a third-party company whose detector is itself an AI model. As the ban is worded around generative tools, and Turnitin insists its detector generates nothing, the two may not technically conflict. When Turnitin launched its AI detector in 2023, Vanderbilt University worried that the tool was unfair to students. The feature was switched on with less than a day’s notice, and universities could not turn it off at first. Turnitin claimed that its false-positive rate was just 1 per cent, which sounded reassuring. Vanderbilt, however, worked out what 1 per cent would mean for its own students. Because the university had sent 75,000 papers through Turnitin in 2022, even that small margin of error could have wrongly flagged around 750 of them as partly written by AI. Consequently, Vanderbilt switched the tool off.
By Turnititn’s own figures, 14.8 per cent of English-language essays submitted between October 2025 and February 2026 were flagged as 80 per cent or more AI-written. In 2023, the figure was 3.3 per cent. The detector also makes mistakes. In July, Turnitin proposed new licence terms that would let it use anonymised student work to improve or develop its AI tools. Following an uproar from universities and students, Southampton University decided not to renew its contract with the company, citing proposed changes “that would have allowed it to use student submissions to train AI services.” Newcastle University has since done the same over the potential use of staff and student data for AI, while Westminster University argued that the plagiarism checker had become largely ineffective now that students can generate work with AI.
But is intellectual property and privacy protection the only problem when it comes to AI? Stanford researchers tested seven detectors on essays by non-native English speakers and found 61.22 per cent were labelled as AI. Detectors tend to read plain, careful, predictable writing as machine-made. We were taught that good writing is clear and correct; now clarity can look like cheating. The students most exposed are often those writing in a second language, many of whom opted to be in universities with global perspectives, hoping that it would be an investment in their own futures. On the contrary, many students seem happy to hand their college struggle to a machine: When Anthropic, the company that makes Claude, interviewed more than 80,000 of its users, many said they valued AI because it was patient, always available and free of judgement. One student in India said their professor teaches 60 people and has little time for questions, whereas AI will answer anything at any hour, “including the dumb ones.” Anthropic itself noted that these same qualities turn AI into a refuge whenever human contact feels too uncomfortable. But if our idolisation of a good student rests on them being flawless, the ones who try hardest to be correct now may be the first to be doubted. If we keep building machines to spare us the discomfort of being seen, trading a teacher’s feedback for a list of errors and calling the case closed, we will slowly teach ourselves to need one another less. And so despite the rumours that AI is training to be a better student than you, tread lightly with your embarrassments, your Why Trinity? Answers, and ask anyway, preferably to someone that does not necessitate a training season.
