Sarcouncil Journal of Economics and Business Management

Abstract: The United States government and nonprofit programs are constantly challenged by fraud, waste, and abuse, which negatively affect fiscal integrity as well as public trust. Traditional audits, which are usually retrospective and manual in nature, find it difficult to identify sophisticated or dynamic fraudulent schemes. This paper discusses the potential use of Artificial Intelligence-based audit analytics to support fraud detection and accountability in federal, state, and nonprofit reporting systems. Drawing on the literature review and case study evidence, the paper analyzes AI-driven audit implementations in both governmental and nonprofit settings, highlighting variations in performance across different systems. The results indicate that although AI improves the detection of anomalies, predictive analytics, and continuous auditing, its efficiency is related to the quality of data, algorithm transparency, and integration of governance. Studies indicate that there are continuing weaknesses in model explainability, inter-agency data standardization, and the training of auditors. Adopting explainable AI frameworks, ethical governance standards, and cross-sector collaboration will address these gaps and enable effective oversight. Finally, AI audit analytics have transformational potential in protecting the resources of taxpayers and donors when applied to a clear, ethically managed, and information-based system that balances between maleficence and prudence in automation.