The global study, led by the University of Ljubljana (Slovenia), analysed student responses collected in three waves between 2023 and 2025. Several scientific publications have reported on the results of the first wave, which gathered more than 23,000 responses from students from 109 countries. The Spanish universities involved in the collaborative research were UIC Barcelona – represented by lecturer Ruggero Colombari, professor Marta Mas-Machuca and professor Frederic Marimon – the University of A Coruña and the University of Alicante, which together contributed 1,421 responses from students across different faculties.

Based on the initial analysis of the results collected between late 2023 and early 2024, 84% of students in Spain reported using ChatGPT. The results of the third wave of responses from the international study were recently finalised and will be presented in forthcoming academic publications linked to the joint project.

One of the main conclusions is that students consider ChatGPT a valuable tool mainly for idea generation, text synthesis and academic writing, and they particularly appreciate its ability to simplify complex information. However, they remain sceptical about its reliability and effectiveness in the classroom, and raise concerns regarding plagiarism, academic fraud and privacy.

Participants rated a series of AI-related questions on a scale from 1 to 5. The analysis shows a clear relationship between frequency of use and satisfaction: students who rarely use ChatGPT rated it at 3.36, moderate users at 3.91, and intensive users at 4.51.

ChatGPT does not improve soft skills

Using the same methodology, the study highlights a clear distinction between students’ positive evaluations of hard‑skill aspects, such as academic writing or foreign-language competence, and their sceptical or neutral perception of soft skills, including creative writing, critical thinking, decision-making and interpersonal communication. The findings indicate that students do not believe ChatGPT helps improve these abilities.

According to Ruggero Colombari, lecturer in the Faculty of Economic and Social Sciences at UIC Barcelona and member of the Observatory on Artificial Intelligence and New Technologies (OIANT), these changes are already visible in university environments. “In group assignments, each student completes their section using ChatGPT, interacting with the tool but not with their peers. They then add their contribution to a shared document. During the process, there has been no exchange of ideas or communication within the group,” he notes. “It is easy for these skills to diminish, which is why lecturers should place greater emphasis on them,” he explains.

The lecturer anticipates a shift in the lecturer’s role in the classroom. “In a context in which AI can provide and synthesise content, the lecturer’s role must be that of a tutor who supports students in developing their soft skills,” he concludes.

Consensus on ethical implications

Globally, students believe that universities and faculties should take into account the ethical implications of using ChatGPT, since it “may encourage academic fraud and plagiarism”. They also expressed concerns about wider social impacts, including privacy risks and increased social isolation.

Despite these concerns, there is general optimism about future employment prospects. The strongest consensus relates to the possibility that AI tools such as ChatGPT could facilitate remote working and reduce workload, improving productivity without threatening future job opportunities.