The AI in Business Podcast Why False Positives Are Costing Banks More Than Fraud - with Suvaleena Paul of Bank of America
11 snips
Nov 25, 2025 Suvaleena Paul, Assistant Vice President and Senior Analyst at Bank of America, dives into the evolving world of fraud prevention. She reveals how synthetic identities and AI-driven tactics are complicating detection. Suvaleena discusses their shift to rapid responses for real-time attacks, enhanced risk scoring techniques using biometrics, and innovative methods to reduce false positives through customer behavior clustering. She also elaborates on utilizing graph theory to unearth fraud networks and emphasizes the importance of tightly governed in-house AI systems.
AI Snips
Chapters
Transcript
Episode notes
Fraud Is Becoming Harder To Distinguish
- Fraud is evolving into “clean fraud” where attacks mimic legitimate behavior and evade legacy systems.
- Detection must operate down to the second and identify subtle mismatches across many signals.
$800K Drained After Biometric Takeover
- A fraud case involved someone adding their fingerprint, suppressing alerts, and draining $800,000 over six months.
- The victim saw no traditional red flags because the activity appeared as legitimate usage.
Adopt A Rapid Response Fraud Model
- Move from quarterly rule reviews to a rapid response model that can deploy countermeasures within hours.
- Monitor hourly signals and convene teams immediately when spikes or pattern changes appear.