Generative AI for Elderly Well-being through the Computer as Social Actor Paradigm

Artificial intelligence and machine learning (AI/ML) technologies like generative AI solutions are proliferating in the real-world healthcare sector. The purpose of this research is to investigate social norms, expectations, and standards of the elderly population for improving trust relationships while interacting with generative AI. The study is based on the CASA paradigm to gain a better understanding of the trust dynamics in human-computer communication to improve the adoption of GAI for elders' health and well-being. We validated the conceptual model with empirical data from 287 elderly users collected through an online and offline survey tool. Quantitative responses received were analysed using structural equation modeling. The study highlights how multimodal interaction, empathy, personalization, augmentation, bias stereotyping, and privacy and security affect the extent to which elderly consumers perceive GAI as trustworthy. Findings indicate that multimodal interaction, personalization, augmentation, and bias stereotyping significantly influenced the trust relationship between the elderly population and GAI. However, empathy privacy, and security were found to be insignificant in trust relationships. Further trust relationships significantly impacted GAI usage. The research provides strong theoretical and practical implications as all the stakeholders like healthcare professionals, patients/users, caregivers, and technology developers can be involved in building applications that cater to diverse needs and promote positive social interactions that can enhance GAI trust and usage.