Preventing Wire Fraud and 2 Interviews From BH USA 2026 From Optiv Security and Kai - Galina Antova, Todd Sorrel, John Hurley - BSW #463
Summary
Business Security Weekly episode 463 features Todd Sorrell of 6lock discussing how AI-enabled vishing attacks are compromising traditional voice verification for wire transfers. The segment explores the need for zero-trust identity verification platforms to secure private market transactions as banks and insurers shift liability away from themselves. The episode also includes pre-recorded interviews from Black Hat USA 2026 with Optiv and Kai, focusing on agentic security operations, AI guardrails, and the transition from reactive to proactive threat detection in enterprise environments.
Topics discussed
Show intro and guest preview
Casual banter on college football
Todd Sorrell background and 6lock overview
AI-enabled voice cloning and social engineering
Reputational risk in private markets
The failure of phone verification and BEC
Dark web data and the need for IDV
Moving away from email and callbacks
Liveness tests and government ID matching
Digital IDs and biometric badging
Creating verified financial networks
Bank liability and the trust gap
Regulatory needs for inter-bank verification
Consumer susceptibility and fintech solutions
AI agents in accounts payable workflows
Bot-to-bot verification and execution
The reckoning of machine-speed attacks
Improving deepfake quality and susceptibility
Customer success story and wrap-up
Black Hat 2026: Optive Security intro
Optive's advise-deploy-operate model
Cloud migration and AI security challenges
Managing AI agents and asset inventory
SOC management and boutique offerings
Deploying correctly and real-time insights
Machine vs machine competition and speed
Proactive vs reactive security metrics
Budget constraints and cost efficiency
Approaching 'help us with AI' requests
AI for internal efficiency vs marketing
Kai Business Security: AI underestimation
Reorganizing defense for machine speed
Consolidating the security stack
Autonomous defense and trust in AI
The reality of spreadsheet reliance
Show the homework: deterministic AI
Human validation in autonomous workflows
Changing roles and superhuman defenders
Exposure management and organizational silos
Breaking silos and machine speed execution
Shifting from alerts to outcomes
Closing remarks and resources
Listen ad-free on Castria