Generative AI Health Assistants in Modern Healthcare: Drivers and Barriers to Adoption

Abstract Generative AI Health Assistants (GAIHAs) are transforming patient engagement in modern healthcare by providing personalized support and medical information. Despite the rapid growth in the number of GAIHAs and the technologies that integrate them, patient adoption remains uncertain. This uncertainty highlights the need for a deeper exploration of the factors influencing their acceptance and adoption as well as the barriers that limit widespread use. Traditional models, like the Technology Acceptance Model (TAM), primarily emphasize adoption drivers such as usefulness and ease of use while overlooking barriers such as privacy concerns and resistance to change. Other models such as the Unified Theory of Acceptance and Use of Technology (UTAUT) include social influence but limit scope to workplace expectations, neglecting broader social dynamics that are relevant in consumer-driven contexts like healthcare. This study investigates the drivers of GAIHA adoption through a new framework, EVF-DOI-IB, which integrates the Extended Valence Framework (EVF), the Diffusion of Innovation (DOI) theory, and Innovation Barriers (IB). The proposed framework builds on the core constructs of trust, risk, benefit, and intention from EVF, and extends them with additional antecedents. The new framework also incorporates innovation drivers from the DOI perspective, namely relative advantage, trialability, and interpersonal communication. Finally, the new framework includes the innovation barriers resistance to change and privacy concerns. The results from a quantitative analysis of our survey data reveal that trust and perceived benefits strongly predict adoption intentions while resistance to change and privacy concerns heighten perceived risks. The findings also show that interpersonal communication and relative advantages play vital roles in reinforcing trust. Our research contributes to the body of knowledge in that it expands EVF’s application to the domain of healthcare AI technologies and provides actionable recommendations for developers and healthcare providers.