The widespread presence of artificial intelligence (AI) in academia raises pressing questions regarding students’ ability to employ these technologies effectively and ethically. This study presents the results of an innovative educational project collaboratively implemented by the Universidad de Las Palmas de Gran Canaria and Universidad de La Laguna (Spain). Groups of students from diverse degree programs participated in a structured learning process comprising three sessions: an introduction to general AI concepts, practical application workshops, and the development of an evaluation rubric.

To assess the impact of this methodology on students’ knowledge and perceptions, pre- and post-intervention questionnaires were administered, each encompassing five sections: demographics, AI experience, currently used tools, concerns and impressions about AI, and issues related to sustainability. Students from the Business Management subject were enquired about the effect of the experience and 103 answers were received for the prequestionnaire and 135 for the post one.

Statistical analyses, including cluster analysis and mean-difference testing, revealed that students’ generally positive outlook on the long-term educational potential of generative AI remained stable throughout the intervention. Notably, as students acquired more knowledge about generative AI, their awareness—and articulation—of its possible impacts increased. However, perceptions of risks associated with generative AI (creativity, dependence, critical thinking, plagiarism, privacy) did not change significantly, suggesting that the training did not alter pre-existing concerns, and that more in-depth activities may be required to effect meaningful change in this area.

Interestingly, students reported spending less time using AI tools post-intervention yet engaged with a broader variety of applications. This indicates a transition from indiscriminate use to more targeted and efficient practices, highlighting an improvement in the quality rather than the frequency of AI use.

Cluster analysis further demonstrated a greater diversification of student profiles after the intervention, with new groups emerging characterised by more conscious, frequent, or strategic use of AI. The variety of tools explored increased, suggesting richer experimentation; nonetheless, substantial segments of students remained low users, indicating only a partial effect—potentially more pronounced among those already predisposed to AI adoption.

These findings underscore the complexity of fostering responsible and effective AI use among students and suggest that more intensive or differentiated educational strategies may be required to maximise the benefits of AI integration in higher education.

Keywords: IAG, business management, innovation in education.