ARTIFICIAL INTELLIGENCE ADOPTION IN ACCOUNTING AND ITS IMPLICATIONS FOR INTERNAL CONTROL QUALITY

Abstract

Background: Artificial intelligence can strengthen monitoring and consistency while also introducing opaque models, data dependency, automation bias, and new access or accountability risks. Aim: The investigation reported here explains how the relationship between the focal practices and internal control quality operates in accounting functions adopting machine learning, document extraction, anomaly detection, forecasting, and generative artificial intelligence. Method: A conceptual narrative synthesis draws on selected scholarly and institutional publications. The discussion is organized around mechanisms, boundary conditions, and practical implications. Results: The synthesis identifies six linked mechanisms: data lineage and validation, human review of material judgments, model governance, segregation of duties, continuous monitoring, incident response and auditability. The study indicates that outcomes depend less on nominal adoption than on operational adoption quality, governance, learning, and fit with local capacity. Conclusion: Decision makers should define the expected outcome, assign responsibility, establish a small set of auditable indicators, and revise the intervention when research material contradicts its assumptions. Contribution: The assessment provides a conditional framework without claiming primary data that were not collected.