Deepfakes, Scams & Trustworthy AI: SAS Fraud Strategy Explained

Fraudsters now use generative AI—deepfake voice/video, hyper-real emails, and scaled scam centers. How do banks keep up? At Sibos 2025, Christen Kirchner, Senior Manager, Fraud & Compliance, SAS Institute shares a multi-layered, explainable AI approach: strong data foundations, open platforms, low-code/no-code rules for investigators, and unsupervised anomaly detection to catch new patterns—with trust and regulatory readiness built in. Highlights - It’s an arms race: criminals use GenAI; banks need evolving, open tech stacks. - Patterns don’t die: old fraud (e.g., check washing) returns—keep & reuse detection logic. - Trust the foundation: clean, governed data → reliable model outcomes. - Explainable AI: decisions investigators can see, justify, and prove. - Low-code enablement: front-line teams can compose rules/patterns without data-science skills. - Anomaly first: unsupervised models surface the unknowns; feed back into the lifecycle. - Beyond buzzwords: match GenAI/agentic AI to the right tasks; measure impact. - Collaboration matters: connect fraud, sanctions, compliance and share insights. Stay close to the conversation - more executive interviews from Sibos, Money20/20 and beyond are on our channel. #Sibos2025 #SAS #Christen Kirchner #FraudPrevention #FinancialCrime #AIinFraud #GenAI #Deepfakes #ExplainableAI #TrustworthyAI #AnomalyDetection #LowCode #NoCode #DataGovernance #Compliance #Sanctions #RiskManagement #BankingTechnology #FinancialIT #Frankfurt