Extending the UTAUT Model for Artificial Intelligence Adoption: A Systematic Review of the Role of Perceived Intelligence and Trust

Abstract

The Unified Theory of Acceptance and Use of Technology (UTAUT) has been widely used to explain technology adoption, although its suitability for complex, learning-oriented systems such as Artificial Intelligence (AI) is contested. This systematic literature review, following PRISMA 2020, reviews 71 empirical and review articles published from 2020 to 2025 that apply UTAUT or its extensions to AI adoption in a variety of domains such as education, healthcare, finance and banking, public administration, and retail and services. The review identifies three common limitations of traditional UTAUT in the AI context: little attention to the opaque “black box” decision-making of AI, little consideration of relational user AI interactions, and little consideration of ethical issues such as algorithmic bias and fairness. To address these shortcomings, it proposes an integrated framework where Perceived Intelligence captures users’ evaluation of an AI system’s learning ability, adaptability, and predictive accuracy. Perceived Intelligence is positioned as an intermediary between core UTAUT constructs, in particular Performance Expectancy and Effort Expectancy, and Behavioural Intention. Trust in AI is positioned as a critical antecedent of Perceived Intelligence and disaggregated into competence-based, transparency-based, and privacy-based dimensions. The review offers a theoretically grounded UTAUT extension for AI adoption, specifies hypothesized structural paths, and examines sectoral variation in ethical and competency concerns, providing a testable model for future structural equation modelling and practical guidance for trust-centred AI implementation.