OPTIMIZING AUDITS WITH AI-DRIVEN RISK MODELS: ENHANCING RESOURCE ALLOCATION AND COMPLIANCE IMPACT

Abstract

AI, the new mega trend is changing risk identification, prioritization and resolution. The study considers The role of or potential for AI based risk models in improving audit planning, use of available resource optimization as well as compliance down the road. Traditional audit models and processes that are based on periodic sampling and reviewing, no longer being enough for the rising complexity and volumes of financial data. This can, on the other hand, lead to knew patterns of behaviour. AI-driven systems have the ability to scour entire data sets for anomalous behaviours and identify unknown threats much more accurately than humans.

It shows that in the novel hybrid AI model which implements human-in-the-loop auditing framework, the existing works utilize supervised learning models for classification and unsupervised learning models for anomaly detection. It uses data from a variety of data sets, including income from reporting bodies, tax returns and external economic factors. The results shows notable improvements over the current audit selection techniques both in terms of revenue preservation, false detections reduction and detection accuracy. The results of the current study indicate that, using their own resources, auditors can obtain more responsible use of AI models through transparency, equality, and governance