Shadow AI usage and ethical AI adoption: the moderating role of chief AI officer governance

Abstract

Purpose The rapid emergence of Shadow artificial intelligence (AI), the informal and unauthorised use of AI tools by employees, poses a growing challenge for global organisations. It creates cybersecurity threats, compliance risks, data breaches and ethical violations such as bias and opacity, ultimately undermining ethical AI adoption and stakeholder trust. Existing research often portrays the Chief AI Officer (CAIO) as a static role, focusing on compliance enforcement or policy integration while overlooking its potential to balance innovation and governance. This study aims to examine how the CAIO as a moderator influences the relationship between Shadow AI Usage and Ethical AI Adoption. Design/methodology/approach A deductive, theory-driven, cross-sectional research design was adopted, drawing on 665 AI practitioners and organisational decision-makers recruited through professional networks across multiple countries. Partial least squares structural equation modelling was employed to test the hypothesised moderating relationship. Findings Higher Shadow AI Usage was negatively associated with Ethical AI Adoption, whereas the positive Shadow AI Usage × CAIO interaction indicates that stronger CAIO governance attenuates this negative relationship. Originality/value This study contributes to AI governance research by conceptualising CAIO governance as an organisational boundary condition that attenuates the negative association between Shadow AI Usage and Ethical AI Adoption.