The impact of AI on financial audit
0 viewsAbstract
Financial fraud is becoming harder to audit because businesses process numbers of digital transactions every day. This research studies how Artificial Intelligence can modernize financial fraud auditing by using machine learning as a tool for auditors. The project uses empirical evidence from the ULB/Worldline credit card fraud dataset and published research .The dataset contains 284,807 transactions and 492 fraudulent transactions showing why accuracy alone is not enough. Results reported by Al Bulushi et al. (2026) Show that machine learning models can have different precision and recall even when accuracy is approximately 99%.XGBoost reported 0.91 precision and 0.79 recall while the Bagging Ensemble reported 0.77 precision and 0.90 recall. In addition, Fedyk et al. (2022) Found that higher recent Artificial Intelligence investment was associated with a 5.0% likelihood of audit restatement and a 0.9% lower audit fee. The findings suggest that Artificial Intelligence can improve fraud screening, risk prioritization and audit efficiency. Artificial Intelligence should not replace the auditor. Human investigation, judgment, model monitoring and governance remain necessary. The research therefore supports a Artificial Intelligence collaboration model, for the modernization of financial fraud auditing.
Keywords
Artificial Intelligence, Audit Quality, Auditing, Finance, Financial Fraud, Fraud Detection, Machine Learning