The double face of artificial intelligence: a study on its effect on academic performance and university creativity

Abstract

This study aims to examine the dual effects of artificial intelligence on university learning, analyzing its influence on academic performance and creativity. The methodology employed was a mixed-methods approach through a survey administered to 160 students at the National University of Costa Rica, Liberia Campus. The results reveal a fundamental paradox: while AI optimizes operational efficiency in tasks such as information retrieval or solving structured exercises, it simultaneously erodes deep cognitive skills. A total of 58.8% of participants reported a loss of critical thinking linked to algorithmic dependency, and 51.9% admitted a decline in their critical thinking skills. In the creative realm, an alarming perceptual gap emerges: although 59.4% denied that AI restricts their innovative exploration, qualitative narratives reveal a standardization of solutions and a reduction in experimentation. This contradiction validates the illusion of preserved creativity, where tangible results mask compromised cognitive processes. Factors such as academic discipline (with a greater negative impact in the humanities), student self-regulation, and teacher training significantly moderate these effects. The conclusions call for urgent pedagogical action: integrating critical AI literacy to identify algorithmic biases and manage dependency; designing hybrid models that combine digital tools with technology-free exercises; and reformulating academic assessment to prioritize creative processes over final products. Ultimately, while AI enhances efficiency, it limits the depth of learning, and its educational value will depend on institutional frameworks that prevent cognitive atrophy.