Since late 2022, when ChatGPT first appeared, the stance of academic institutions has shifted significantly: from blanket bans and fears of widespread cheating, to efforts to integrate AI tools into the learning process. At the same time, however, a growing number of universities are reconsidering a return to supervised, in-person exams, viewing them as a safer way to safeguard academic integrity, according to Times Higher Education.
"According to recent surveys, around 95% of students say they use AI tools during their studies, a figure significantly higher than the 66% recorded just two years earlier."
Integration
In 2023, shortly after the emergence of large language models, around a third of Russell Group universities in Britain had adopted strict restrictions, expressing concerns that artificial intelligence would undermine the quality of learning and the integrity of exams. Despite the bans, student use of AI increased dramatically. "According to recent surveys, around 95% of students say they use AI tools during their studies, a figure significantly higher than the 66% recorded just two years earlier."
"The academic community appears divided, as it remains unclear to what extent artificial intelligence can be used in exams and assignments without distorting the true picture of a student's knowledge and skills."
At the same time, a growing number of students are incorporating AI-generated text into the assignments they submit for assessment. This reality has led many institutions towards a different approach. A number of universities now recognise that AI is a tool that will accompany graduates throughout their future careers. As a result, they permit its use in research, study, organising ideas, or even in preparing coursework. Some have gone even further, forming partnerships with AI development companies to give students access to the most advanced applications. The rationale behind this approach is that proficiency in AI tools is now a core professional skill, comparable to the use of computers or specialised software.
The concerns
"The academic community appears divided, as it remains unclear to what extent artificial intelligence can be used in exams and assignments without distorting the true picture of a student's knowledge and skills." Despite differing approaches, most universities have settled on an intermediate management system known as the "traffic light" model: the "red" category fully prohibits any use of AI tools; the "amber" category allows limited use under specific conditions; while the "green" category not only permits but often encourages the use of AI as part of the learning process. This model is already applied at a large number of universities in Britain. At the same time, however, even institutions that maintain this system acknowledge that it is not sufficient on its own to address the challenges created by AI.
Education or surveillance?
Regardless of individual regulations, a report by the Higher Education Policy Institute (HEPI) analysed the AI policies applied by British universities. It found that of the UK's 163 universities, 96 have a publicly accessible policy on the use of AI, with most placing emphasis on skills development, critical thinking and the responsible use of the technology. Only four institutions have adopted a purely deterrent approach, focusing mainly on violations, penalties and detection systems.
Even so, the report's authors point to a significant contradiction. Although many policies appear educational in theory, in practice they function more as mechanisms of compliance and control. The University of Southampton is cited as an example of this contradiction: although it states that its goal is to produce graduates with strong digital literacy, its policy is accompanied by extensive lists of permitted and prohibited uses of AI, and even provides for disciplinary sanctions for violations.
What Cyprus's rectors say
Panayiotis Zafeiris, Cyprus University of Technology (CUT)
"Artificial intelligence will not necessarily eliminate specialisations, but it will completely change the way study programmes are taught and delivered. The model where students sit in class, listen to the professor and memorise the material cannot stay the same. We need to give them greater freedom to explore, to use tools, but also to be able to explain, check and critically defend the outcome of their work."
Filippos Pougiouta, University of Nicosia
"Artificial intelligence will change the workplace, the way we live, the way we communicate, and also education. That is why universities must revise their study programmes based on technological developments and give students the knowledge and skills they need. The use of AI tools must be governed by rules and academic ethics."
Andreas Efstathiou, European University Cyprus
"It can help a great deal in education. It can enable personalised learning and change the way we teach and learn. At the same time, however, it raises questions about the career prospects of many graduates. That is why education must cultivate even further what a machine cannot offer: creativity, critical thinking, curiosity, and problem-solving ability."
Pantelis Sklias, Neapolis University Pafos
"For me, artificial intelligence spreads knowledge much faster than before. But we are still the ones generating the input of information. So the challenge is to produce even better scientists. The more substantive knowledge you have in your field, the better you can put artificial intelligence to use for your own benefit and for the benefit of society."
