Opinion Panel | Because of AI, the thesis is an outdated form of assessment
Curse or blessing: artificial intelligence is becoming increasingly embedded throughout society. Universities, too, are struggling with AI's rapid arrival in the lecture hall. After all, which student still bothers to write their own texts these days? This time, the brand-new Folia opinion panel is turning its attention to the thesis as an “outdated form of assessment”.
Because of AI, the thesis is an outdated form of assessment
Cor Zonneveld (65) is biologist and teachter at the AUC.
“AI has not made the thesis outdated. It has revealed that a thesis alone was never sufficient to establish what a student knows and can do. Even BC – Before ChatGPT – a thesis was not produced in isolation. Students received supervisor guidance, peer comments, language or statistical help; and sometimes less legitimate assistance. Yet we treated the submitted text as transparent evidence of individual competence. AI has made that assumption untenable.”
“This does not undermine the thesis’s rationale. Scholars must still formulate questions, engage critically with literature, choose methods, analyse material and construct arguments – now with AI available to assist them. Writing a thesis is therefore not obsolete; it reflects the conditions under which future academic work will occur.”
“What may be outdated is the thesis as an unaudited, stand-alone product. Assessment must also reveal the student’s own work through supervision, intermediate products, transparency about AI use and an oral defence. AI does not invalidate the thesis; it finally forces us to do what we should have done all along: establish whether the student understands the research, can justify its choices, and demonstrates genuine intellectual ownership of the result.”
Silvia Bușcan (19) studies media & culture.
“We might be tempted to think that a thesis no longer reflects a student’s knowledge, because almost any assessment that requires writing a text, conducting research or gathering data can be done by AI. But that is not necessarily true.”
“There are two different assumptions underlying this debate. One side assumes that the use of AI can be kept under control by detecting and sanctioning it, while the other side assumes that AI has evolved so much that no one can tell for sure whether it is being used. But even if we accept the latter assumption, I would still argue that AI has nothing to do with the value of a thesis as students could pay others to do their work long before this tool existed. In other words, AI is just another form of cheating.”
“But ensuring that its students do not cheat is not the university’s responsibility, considering that the only person harmed by this is the student themselves. Anyone can decide for themselves whether it is better to cheat and learn nothing or be honest and prepared for future challenges.”
Ștefan Brădianu (21) studies PPLE.
“While AI has certainly changed much about research, it has not changed how science is systematised. The single most logical way of grouping knowledge is into hypotheses, to be proven or disproven one at a time – through theses. For now, there is still a need for humans to contribute to proving these hypotheses.”
“It is true that AI has become a useful, almost indispensable tool in this process. However, the thesis accounts automatically for these evolutions. If nowadays AI skills are crucial in writing a thesis, very well. It means that those AI skills will be (indirectly) measured when the thesis overall is evaluated.”
“Therefore, the thesis is, in fact, a perfect test, because, by being so fundamental to the academic process, it does not need to be constantly updated to stay relevant. By its very nature, it evolves with the times.”
Would you like to join the Folia opinion panel too? Then click here for more information.
