Exploring the Effect of Generative AI on Social Sustainability Through Integrating AI Attributes, TPB, and T-EESST: A Deep Learning-Based Hybrid SEM-ANN Approach

Abstract

The swift progress of generative artificial intelligence (AI) tools offers remarkable potential for revolutionizing educational methods and enhancing social sustainability. Despite its potential, understanding the factors driving its adoption and how that affects social sustainability remains underexplored. This study aims to address this gap by integrating AI attributes (“perceived anthropomorphism,” “perceived intelligence,” and “perceived animacy”) with the theory of planned behavior and the technology-environmental, economic, and social sustainability theory (T-EESST) to develop a theoretical research model. Utilizing a hybrid structural equation modeling and artificial neural network approach, we analyzed data collected from 1048 university students to evaluate the developed model. Our findings revealed that while perceived behavioral control has an insignificant impact on generative AI use, attitudes emerge as the most critical factor, further reinforced by the significant role of subjective norms. Perceived anthropomorphism, perceived intelligence, and perceived animacy were also found to influence students’ attitudes significantly. More importantly, the findings supported the role of generative AI in positively affecting social sustainability, aligning with the principles of T-EESST. This study's significance lies in its holistic examination of the interplay between technological attributes, motivational aspects, and sustainability outcomes, offering valuable insights for various stakeholders.