Internal surveys from the University of Málaga for the 2025/2026 academic year reveal that nine out of ten assignments submitted by students show evidence of AI usage. The institution is preparing specific regulations for the upcoming academic year.
The University of Málaga (UMA) has found that 90% of the assignments submitted by its students during the 2025/2026 academic year show clear evidence of having been produced with artificial intelligence. This data comes from internal surveys conducted by the institution itself, as reported by Diario Sur, and reflects that the use of these tools is no longer sporadic but a widespread practice across nearly all faculties.
This phenomenon has led UMA to review the validity of its evaluation methods. Working groups and internal analyses have resulted in the creation of an AI portal that will launch this academic year and will become the first regulatory framework for these technologies at the university. The document includes guidelines on ethics, privacy, data protection, and authorship policies.
“Almost all assignments show traces of AI,” says Joaquín Canca, the Rector's delegate for Information Technology and Digital Strategy at UMA, who has participated in designing this framework. Canca warns that traditional anti-plagiarism software has become obsolete: “It was working well, but with AI that is no longer the case,” because generative technology “creates new text that has never been written in that way before.”
The new model includes modifications to the evaluation regulations. “If there is suspicion that an assignment has been produced with AI, it must be agreed upon. If it is used, it must be disclosed, just as has been done with bibliographies,” explains the delegate, who adds that the institution plans to conduct “an authorship test with the student” when there are doubts about the originality of an assignment.
A specific case illustrates the problem: a student from the Faculty of Philosophy and Letters submitted an assignment that included conclusions and points extracted from a scientific study previously published by the very professor who graded it. The AI accessed the content, and the student claimed authorship.
The dean of Philosophy and Letters, Sara Robles, has already warned of this new reality in an interview: “The concept of assignments or final projects is shaken by artificial intelligence.” Robles believes that the university must rethink its evaluation systems, which are heavily based on exams and written assignments.
The Digital Strategy head advocates for a change in approach: “Evaluation must shift from the result to the process. It has no value for you to submit the assignment; we need to evaluate the process and confirm authorship.” As a result, he predicts an increase in oral exams, although he acknowledges that this format requires more time and poses logistical challenges.
Canca also warns of the pedagogical risk of abusive use: “When I hear that one of the values of AI is to confront a blank page, I think that a student with a blank page is when they learn the most.” In his view, systematically resorting to these tools may lead students to learn less.
The impact is not limited to final degree or master’s projects. Law professor Elsa Álvarez points out that usage has become widespread across all types of subjects, although it was initially thought to be concentrated in final projects.
UMA is already working on the approval of specific regulations that will be ready for the next academic year and will detail how to act in cases of authorship conflicts.
You may also like
