Unmasking Biases and Mapping the Landscape of AI Adoption in Diverse Business Functions Through a Delphi Study

Abstract

Artificial Intelligence (AI) tools are exceedingly being introduced in various business sectors as a way to improve efficiency and drive overall organisational performance. Prior research has uncovered many success and failure factors influencing the adoption of these tools. However, in the absence of a common understanding between practitioners and researchers, factors deemed theoretically significant do not always align with reality, resulting in a researcher bias in AI adoption literature. Additionally, these factors and their priorities depend on specific business functions, deeming existing one-size-fits-all AI adoption theories incapable of explaining these nuances. To address these shortcomings, this study investigates the existence of a potential researcher bias and establishes factors influencing AI adoption in different business functions through a 2-fold, 3-round, 3-panel Delphi study. The findings establish a potential researcher bias and confirm that factors influencing adoption, and their priorities, differ by business functions. This study contributes to literature by first establishing the potential researcher bias and then furthering the understanding of factors influencing adoption for different business contexts. In a pivotal contribution to practice, this study enables organisations to foster better adoption practices based on different business functions.