The rapid growth of digital banking and FinTech services has increased the risk of cybercrime and online financial fraud. Traditional fraud detection systems often struggle to manage high transaction volumes and sophisticated fraud patterns in real time. This study develops a conceptual framework for real-time fraud detection in digital banking using predictive and prescriptive analytics. The research adopts a positivist philosophy and deductive approach based on secondary data sources. The proposed framework integrates machine learning, artificial intelligence, and automated decision-support mechanisms to improve fraud detection accuracy, response speed, and cybersecurity resilience. The study highlights the importance of intelligent analytics systems in strengthening digital banking security and fraud prevention,