The expansion of digital banking, online transactions, and financial technology services has provided convenienceandaccessibility to millions of users worldwide. At the same time, these advancements have significantly increasedtheriskoffraudulent activities. Conventional fraud detection systems rely primarily on predefined rules and static patterns, whichareincapable of adapting to rapidly changing fraud strategies. As a result, many genuine cases of fraud remain undetected, whilefalse positives create inconvenience for legitimate users. Additionally, centralized storage of fraud detectionlogslackstransparency and integrity, making audit trails vulnerable to tampering and disputes during regulatory or legal reviews. This project proposes the design and implementation of an AI-driven fraud detection system integrated with a customblockchainframework, TrustLedger. The AI module employs machine learning algorithms to identify suspicious transactions inreal timebyanalyzing transaction history, behavioral patterns, and anomalies. In parallel, the TrustLedger blockchain frameworkmaintainsan immutable and tamper-proof record of flagged transactions, ensuring that every detection decision is verifiable andresistanttomanipulation. To further enhance transparency, the system incorporates explainable AI methods, allowing investigatorstounderstand the reasoning behind each fraud alert. The outcome of this project will be a prototype platform capable of real-time monitoring, secure audit logging, and comprehensivereporting. By addressing both detection accuracy and trust in audit evidence, the system aims to deliver a reliable andacademicallyrelevant solution for financial fraud prevention, aligning with modern requirements of security, compliance, and user confidence.KeywordsArtificial Intelligence, Fraud Detection, Machine Learning, Anomaly Detection, Blockchain, TrustLedger, Immutable Audit Trail,Explainable AI, Secure Transactions, FinTech Security