This paper examines the integration of Artificial Intelligence (AI) into audit and control systems through a cybernetic perspective. It conceptualizes AI-enabled auditing as a socio-technical control architecture structured around feedback mechanisms, governance layers, and adaptive regulation, and analyzes its implications for institutional performance.
The study employs a structured scoping-review methodology based on peer-reviewed literature (2020–2025) and authoritative professional sources. Through thematic coding and conceptual synthesis, it develops a systems-oriented framework linking AI capabilities to audit-cycle stages, feedback dynamics, governance structures, and professional judgment.
The analysis identifies key mediating mechanisms through which AI may influence institutional performance, including expanded evidence coverage, enhanced anomaly detection, and reduced detection latency. Performance effects, however, depend on governance robustness, data quality, and auditor digital competencies. AI-enabled audit systems operate as layered socio-technical architectures in which algorithmic processing and human oversight perform complementary functions, reflecting a broader reconfiguration of the cybernetic architecture of control and audit systems across inputs, processing layers, feedback mechanisms, and governance structures.
The paper contributes to the literature by advancing a cybernetic interpretation of AI-enabled auditing, integrating control theory, governance architecture, and professional judgment within a unified conceptual model. It extends discussions on digital transformation by framing audit and control systems as adaptive feedback-based regulatory structures.
