•  
  •  
 
Communications of the Association for Information Systems

Author ORCID Identifier

Subhendu Dey: 0009-0001-1808-5212

Rohan Mukherjee: 0000-0003-1888-7204

Sachin Modgil: 0000-0001-7816-6594

Ajoy K. Ray: 0009-0003-3538-9992

Surajit Bag: 0000-0002-2344-9551

Abstract

Generative AI (Gen-AI) is being absorbed rapidly into IT services and operations (IT S&O), but its adoption is not well understood yet. The established adoption models in information Science (IS) were built for stable, deterministic IT artefacts, whereas Gen-AI is probabilistic, co-agentic, and emergent; and the enablers and barriers that shape its uptake in IT S&O remain scattered across a young literature. This study explores enablers and barriers; and investigates how dominant theories must change Gen-AI related dynamics. We screened the literature through a three-phase protocol and analysed 61 peer-reviewed studies published between 2023 and mid-2025; and organised them by the functional capabilities of Gen-AI: generation, classification, reasoning, and optimisation. The synthesis surfaces a clear asymmetry. Enablers are capability-specific, varying with the function exercised, while barriers are quality-generic, recurring across limitations such as verification failure, hallucination, and environmental cost. Auditing this evidence against TAM, UTAUT2, and IDT shows that the gap is not a missing variable but a boundary-condition in four assumptions these theories rely upon. We propose the Hybrid Gen-AI Adoption Framework (HyGAAF): a five-layer, eleven-proposition extension of UTAUT2 built around a dual affordance that pairs creation with verification, calibrated trustworthiness, and institutional and governance boundary conditions.

Download

Share

COinS
 

When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.