<![CDATA[Micro, small, and medium enterprises increasingly process transactions across digital channels, yet many still struggle to convert daily records into reliable accounting information for monitoring and control. This study addresses a practical problem: fraud detection solutions using artificial intelligence are often not adopted, not because technology is unavailable, but because accounting data are fragmented, inconsistent, incomplete, and difficult to trace. The study aims to examine how data friction influences trust in artificial intelligence and adoption readiness for fraud detection, and whether data governance maturity reduces the negative effect of data friction on trust. A cross-sectional survey was administered to enterprise owners or managers responsible for transaction recording and data handling. The proposed model tests a moderated mediation mechanism in which trust links data friction to adoption readiness, while data governance maturity buffers the adverse pathway. The novelty lies in positioning accounting data friction as the central barrier to adoption readiness, explaining adoption through trust formation, and highlighting governance maturity as a practical form of internal control discipline that strengthens confidence in data-driven oversight. The results show that data friction lowers trust and adoption readiness, trust increases adoption readiness, data governance maturity strengthens trust, and the indirect effect through trust remains significant while being moderated by governance maturity. The study concludes that improving basic accounting data discipline and governance is essential to make fraud detection solutions more trustworthy and adoptable, and future research should validate these relationships using longitudinal designs and objective indicators of accounting process quality.]]>
