Generative AI and educational sustainability: Examining the role of knowledge management factors and AI attributes using a deep learning-based hybrid SEM-ANN approach

Abstract

Integrating Generative AI into educational settings holds transformative potential by personalizing learning, enhancing accessibility, and reducing resource usage, thereby promoting educational sustainability. However, understanding the drivers influencing Generative AI use and its subsequent impact on educational sustainability is still in short supply. Therefore, we developed an integrated model of knowledge management (KM) factors and AI attributes to examine their impact on Generative AI use and its consequent effect on educational sustainability. The model was then evaluated using a deep learning-based hybrid SEM-ANN approach based on data collected from 464 students. The PLS-SEM findings supported the role of knowledge acquisition, knowledge application, perceived anthropomorphism, perceived animacy, and perceived intelligence in positively affecting Generative AI use. In contrast, knowledge sharing showed no notable effect. The findings also showed that using Generative AI significantly promotes educational sustainability. The ANN results indicated that perceived anthropomorphism is the most critical factor impacting Generative AI use, with a normalized importance of 91.10 %. Theoretically, the findings offer empirical evidence on how KM factors and AI attributes influence Generative AI use and its role in enhancing educational sustainability. Practically, this research provides implications for various stakeholders interested in applying Generative AI for educational purposes.