1Mrs. L. SHIRISHA, 2D. SHIVANI, 3B. VIVEK, 4K. RAGHUNATH 03 05 Money laundering is a critical financial crime thatposes serious threats to economic stability andglobal financial systems With the rapid growth ofdigital banking online transactions and financialtechnologies detecting suspicious financialactivities has become increasingly complexTraditional AntiMoney Laundering AMLsystems primarily rely on rulebased approacheswhich often fail to detect advanced launderingtechniques and generate a high number of falsepositives To address these limitations this projectproposes FinShield an AIdriven system designedto detect money laundering activities within socialtransaction networks The system integratesMachine Learning ML behavioural analysis andnetworkbased graph analysis to identify suspicioustransaction patterns and hidden relationshipsamong users Instead of analysing transactionsindividually FinShield evaluates the overalltransaction ecosystem by modelling users as nodesand transactions as edges enabling detection ofcircular transactions layering patterns andsuspicious clusters Multiple ML models such asRandom Forest Logistic Regression and SupportVector Machines are used to classify transactionsbased on risk scores The system also incorporatesrealtime monitoring and a webbased dashboardfor visualization and alerts By combining AItechniques with network intelligence FinShieldsignificantly improves detection accuracy reducesfalse positives and enhances decisionmaking forfinancial institutions The proposed systemprovides a scalable adaptive and efficient solutionfor modern AML challenges ensuring betterfinancial security and fraud prevention in digitalecosystems Copyright (c) Journal of Science Engineering Technology and Management Science. All rights reserved 2026