The impact of AI tools on undergraduate students’ academic performance: A case study of Peru

Abstract

This study examines the impact of ChatGPT as a study tool on undergraduate academic performance in Peru, aiming to identify the effect of specific usage patterns. Considering the increasing adoption of AI in education, assessing ChatGPT’s potential to enhance learning experiences is imperative. Using a structural equation model, this research analysed survey data from 263 students, examining the relationship between ChatGPT usage patterns—such as idea generation, research support, and writing assistance—and academic performance, in addition to the impact of students’ characteristics, university support, and professors’ skills. The findings indicate that students’ skills and stress levels are significant predictors of performance. ChatGPT’s idea generation feature positively influenced academic outcomes by fostering creativity, whereas research support had a negative impact, likely due to outdated information, emphasising the importance of responsible AI usage. These findings offer practical guidance for students, educators, and policymakers on effectively integrating AI tools into education. Implications will be discussed, focusing on practical strategies for maximising ChatGPT’s benefits and addressing its limitations.