Generative AI has already been sitting in lecture halls for over three years now, and the higher education sector has finally stopped treating it as a novelty and started treating it as what it is: a governance problem touching on everything from assessment integrity to data privacy to staff training. The response, however, has been less a coordinated national effort and more a relay race where nobody agrees on the route.
Take the two oldest rivals. Trinity College Dublin has a college-wide statement, approved by its council and board in November 2024, under which AI use is permitted unless a course handbook says otherwise, provided you acknowledge it. Pass off the machine’s prose as your own and it’s plagiarism, full stop. UCD, meanwhile, has no university-wide student policy at all. Rules are set at the level of the school, the programme, or the individual module, which means in practice that a student’s obligations could change between Tuesday’s lecture and Wednesday’s tutorial. (UCD’s College of Arts & Humanities has at least made this legible with a traffic light system: red means no AI, amber means some, green means go, within integrity rules. Simple but it works.)
The rest of the sector falls somewhere along this spectrum. DCU and TU Dublin have formal positions routing AI use through academic integrity and responsible-use rules. Galway and Limerick lean on integrity-led and assessment-specific guidance. Maynooth and SETU appear to handle things locally, with no clear central public policy to be found. The practical upshot: what counts as diligent study on one campus could be classed as misconduct on another. Students face a postcode lottery, and the postcode in question is their CAO form.
Welcome, guardrails
Help of sorts arrived in December 2025 when the Higher Education Authority published its national policy framework on generative AI in teaching and learning. It doesn’t prescribe uniform rules (because institutional autonomy is sacrosanct) but it does establish shared principles for values-based adoption, along with ethical guidelines and AI literacy resources. I’ll admit a flicker of professional pride here: the framework was led by Dr James O’Sullivan, a digital humanities academic from UCC. When the machines came for the college essay, it was the humanities people who got asked to write the rulebook. Funny, that.
The framework is a sensible first step, and the HEA is right that no single actor can sort this alone. But principles are not policies, the last thing a student staring at an assignment brief at 11pm needs is a five-page statement of values; she needs to know whether the thing on her laptop is a study aid or an offence. Until the framework is translated into clear rules on every campus, students will keep guessing and this isn’t ideal when you’re working towards your degree.
The technology is already embedded in how students work. The rules should catch up before the appeals do.
Ltd
