A Multianalytical SEM-ANN Approach to Investigate the Social Sustainability of AI Chatbots Based on Cybersecurity and Protection Motivation Theory

With a primary focus on cybersecurity risks, this study endeavors to explore the sustainable deployment of artificial intelligence (AI) chatbots and, ultimately, to promote their social sustainability. The study introduces an enhanced model built upon the "Protection Motivation Theory" (PMT) to explore the factors that predict the social sustainability of AI chatbots. The proposed model is evaluated using both "structural equation modeling" and "artificial neural network" (ANN) analyses, leveraging data obtained from 1741 participants. The findings reveal that PMT factors significantly predict the sustainable use of AI chatbots. Moreover, cybersecurity concerns, including confidentiality and privacy, have emerged as significant predictors of sustainable use, impacting the social sustainability of AI chatbots. The indicated paths in the model explain 70% and 74% of the variance in sustainable use and social sustainability, respectively. The results from the ANN analysis also emphasize the critical role of confidentiality as the primary predictor. The significance of this study lies in the development of a unified model that integrates cybersecurity and PMT, offering a distinctive framework. In addition to its theoretical contributions, the study offers practical insights for service providers, application developers, and decision-makers in the field, thereby influencing the future of AI chatbots.