Impact of the uncontrolled use of artificial intelligence in virtual assessments on academic outcomes

Abstract

This study analyzes the impact of the uncontrolled use of artificial intelligence (AI) tools in virtual assessments and their potential implications for the development of logical and critical thinking in university students. In a context where online assessments have become increasingly relevant, there is concern that open access to AI may influence academic outcomes, generating grades that do not necessarily reflect actual learning. The research adopted a quantitative, non-experimental, and comparative approach, with a sample of 193 students studying Linear Algebra in the Information Technology program at the University of Guayaquil, distributed across six parallel programs with in-person assessments and one with virtual assessment. Final grades were used as a data source, and descriptive and inferential analyses were performed. The results showed no statistically significant differences between the grades of both groups; however, greater dispersion was evident in the virtual group, with extreme concentrations of grades (0 and 10), suggesting the possible influence of AI during the assessment process. This finding opens the door to reflection on the validity of the results and, especially, on the risks that the indiscriminate use of these tools may limit the development of higher cognitive skills, such as logical reasoning and critical thinking. It is concluded that institutions must implement assessment strategies that regulate the use of AI, ensuring both academic integrity and the development of essential skills for students' professional futures.