Why Artificial Intelligence Challenges the Foundations of Technology Acceptance Models

Abstract

Despite decades of refinement, technology acceptance models such as the Technology Acceptance Model (TAM; Davis, 1985, 1989) and the Unified Theory of Acceptance and Use of Technology (UTAUT; Venkatesh et al., 2003) remain the dominant frameworks for evaluating digital technologies. Their resilience reflects robustness and parsimony. Yet Artificial Intelli-gence (AI) changes the game. Unlike earlier systems, AI learns, adapts and acts, increasingly participating in the decisions, challenging the very assumptions on which TAM/UTAUT rest. As Venkatesh himself admitted, the acceptance of AI tools remains “a question mark”, raising doubts on the adequacy of established models (Venkatesh, 2022). Drawing on a semi-systematic literature review (12,048 publications from 1985–2025, including 155 focused on AI ac-ceptance), we show that while TAM/UTAUT still account for nearly 70% of studies, the field has entered a phase of conceptual displacement. Three converging dynamics stand out: an affec-tive and experiential turn, a vulnerability-centered perspective and a socio-technical orientation. Together, they crystallize into three new research streams: trust-centered, adoption-oriented and ethics-centered, that shift the field away from individual-utilitarian framings toward relational, organizational and governance logics. The challenge ahead is clear: to decide whether constructs such as trust, affect, privacy, ethics and anthropomorphism are merely contextual moderators or the building blocks of a new paradigm. The age of AI calls for more than incremental refine-ments, it demands a shared theoretical framework capable of steering organizations and societies through both the promises and risks of intelligent systems.