ENHANCING FISCAL ACCOUNTABILITY AND AUDITABILITY: A FRAMEWORK FOR DEPLOYING GENERATIVE AI PROCESS AGENTS IN PUBLIC SECTOR FINANCIAL ERPS

Abstract

Fiscal transparency, accountability, and the reliability of audits are some of the issues that public sector organizations are now facing. The paper at hand gives a framework that deals comprehensively with the use of Generative AI (GenAI) process agents in Financial Enterprise Resource Planning (ERP) systems in the public sector. The ERP changes will include the detection of anomalies, auditing process made easier, and reporting of compliance improved. The research is based on mixed methods where quantitative ERP transaction data analysis (p < 0.005) and qualitative interviews are done with financial administrators. The final results show that the use of AI and automation brought about significant improvements in terms of error rate reduction, audit completion time, and compliance accuracy. The framework is built around XAI models, LLMs, and federated learning that are sure to be morally deployed under the government data governance policies. The dummy ERP dataset gives empirical proof that GenAI-driven agents can account for 22% more fiscal accountability and audit efficiency 35% more. The suggested model gives a bright future for the invigoration of transparency in electronic governance systems and the sustainable management of public finances.