Abstract:

This study explores the potential of Artificial Intelligence (AI)–driven procurement monitoring to enhance transparency, accountability, and fraud detection in public procurement systems of developing democracies. Through a comparative analysis of global AI practices and emerging digital oversight models, the research focuses on machine learning techniques, including anomaly detection and logistic regression implemented in TensorFlow 2.0, with data stored and managed in MySQL. The findings suggest that AI can quickly identify risks, optimize audit resource allocation, and improve public transparency through data-driven risk scoring. However, significant challenges remain, including poor data quality, algorithmic bias, limited institutional capacity, political resistance, and concerns regarding explainability. Effective adoption requires complementary reforms in data governance, capacity building, open data standards, and the integration of AI outputs with human oversight. Future research should explore context-specific strategies for AI implementation, assess its empirical impact on reducing corruption, and develop methods to mitigate algorithmic bias while ensuring transparency and accountability in developing democracies.