Sherlock and Watson Fight Deep Social Engineering
summary
Multiple case studies show criminals trick victims into moving their own money: remote-access scams using TeamViewer, real-time voice scams to harvest OTPs and enroll victims in Zelle, malware-driven fake sites (Dridex) that farm challenges, a $2M corporate transfer and a $25K order with $10K fake shipping to Tuvalu. Detection relies on aggregating weak signals — slowed typing, long pauses while dictating, jittery mouse paths, phone-tilt accelerometer traces — into an ML model using behavioral biometrics and hundreds of micro-behaviors, which UK banks already deploy with measurable fraud reduction.
AI-written summary · about
chapters (14)
- 0:00Intro
- 0:26Sherlock Holmes
- 0:51The Redheaded League
- 1:27Jabez
- 3:09Spaulding
- 10:06Remote Access
- 11:15Fraud Team
- 12:58Credit Union Attack
- 15:38User Session Journey
- 20:17The Straw Company
- 26:58Detecting Fraud
- 29:42Being Guided
- 30:45Analysis
- 36:39Recommendations
related talks
40mBanking Cybersecurity: Battling DeepFakes & AI-powered Scammers
20mSecurity in the Age of AI: Response and Strategy
48mHow to Rob a Bank over the Phone - Lessons Learned from an Actual Social Engineering Engagement
55mSocial Engineering: Hacking Humans
25mHacking the human mind: The rise of social engineering security threat
43mBypassing Malware Detection Mechanisms in Online Banking
27mCybersecurity During Dark Times
51mSocial Engineering: Hacking People