Perceived Value and Adoption Intention of Enterprise AI Solutions in Bangladesh: Integrating the Technology Acceptance Model and Diffusion of Innovations Theory

Abstract

Enterprise artificial intelligence (AI) adoption is increasingly framed as a value-realisation problem rather than a simple question of whether users are willing to try new technologies. This study examines perceived value and adoption intention for enterprise AI/cloud solutions among technology-oriented respondents in Bangladesh by integrating the Technology Acceptance Model (TAM) with Diffusion of Innovations (DOI) theory. Distinct from a conventional predictor-only adoption model, the study specifies a value-realization mechanism in which relative advantage and compatibility shape perceived usefulness, which in turn contributes to adoption intention; direct DOI-to-intention paths are retained to test partial versus full mediation. Innovation adoption readiness is additionally examined as a boundary condition on the usefulness-intention relationship. The supplied database contains 385 complete observations after exclusion of 15 blank rows. Respondents are distributed across students, employed professionals, and self-employed/freelance users, with substantial variation in technology-use frequency and experience. Composite-score path analysis shows that relative advantage and compatibility are positively associated with perceived usefulness (β=.204 and β=.217, respectively; both p<.001). Perceived usefulness is positively associated with adoption intention (β=.119, p=.020), while relative advantage (β=.147, p=.005), compatibility (β=.156, p=.003), and innovation adoption readiness (β=.150, p=.004) also have significant direct associations with adoption intention. Bootstrap indirect effects through perceived usefulness are positive for both relative advantage (ab=.025, 95% CI [.003, .052]) and compatibility (ab=.026, 95% CI [.004, .053]). The usefulness × readiness interaction is not statistically significant (β=.039, p=.447). The findings therefore indicate a partial value-realisation pathway rather than a simple direct-determinants model. The article contributes by linking innovation attributes to the TAM usefulness mechanism in a Bangladesh enterprise-AI context and by showing that readiness operates more clearly as a direct adoption resource than as a moderator of usefulness.