SIM swap fraud has become a serious problem now a days in current digital world, especially in India. Criminals exploit weaknesses in mobile carriers’ SIM replacement procedure to sneak into people’s phone numbers. This gives them access to over personal data. This research aims on using artificial intelligence and machine learning to detect and prevent fraud by processing telecom call records for unusual patterns, such as sudden changes in financial data, which often leads to major loss for victims and financial data. The research aims to use artificial intelligence to detect fraud by analysing telecom call recordings for weird patterns. The system has steps like filtering data, finding anomalies, setting thresholds for normal behaviour, and validating results to ensure accuracy. This paper shows gaps in current security measures showing need for strong security measures. Call duration, cost, or revenue revenue.call duration, duration, or cost. By introducing an AI-powered system which is capable of real-time fraud detection this study focus on improving fraud prevention more effectively, while improving security protocols, and creating a framework and better collaboration among telecom sector and safer digital environment for everyone.