AI chatbots: A disguised enemy for academic integrity?

Abstract

The widespread popularity of ChatGPT and other AI chatbots has sparked debate within the scientific community, particularly regarding their impact on academic integrity among students. While several studies have examined AI's role in education, a significant gap remains concerning how AI chatbot usage affects students’ perceptions of academic integrity. This study aims to address this gap through rigorous quantitative techniques to explore the dynamics of student interactions with AI chatbots and assess whether this engagement diminishes academic integrity in higher education. Using a non-experimental design, the research investigates the causal relationship between AI chatbot usage and academic integrity, focusing on eight latent variables identified in the literature. A stratified sampling technique was employed to collect a representative sample of 594 participants via a 5-point Likert scale survey from four Southern Asian countries. The dataset underwent extensive statistical analysis using Structural Equation Modeling (SEM) techniques. The findings establish significant links between motivations for using AI chatbots and a decline in academic integrity. The study identifies a behavioral link between academic integrity and pedagogical limitations, highlighting traditional classroom-based pedagogy as the most impactful factor influencing students’ motivation to engage with AI chatbots. This research not only quantitatively addresses ethical concerns related to AI in academia but also offers insights into user behavior by assigning distinct weights to post-usage behavioral factors, differentiating it from earlier studies that treated these factors equally.