Roger A Grimes, KnowBe4 | RSAC 2026
summary
KnowBe4 outlines a shift from security awareness training to workforce trust management that treats people plus their desktop agents as a single risk surface. A 12-agent AI orchestrator automates and hyper‑personalizes simulated phishing, generating deepfake executive simulations and escalating difficulty automatically; early data shows AI-managed programs cut security risk about 3× versus human admins. The talk cites attackers getting 4.5× more value from AI scams and warns of both attacks from AI and data/context poisoning of defenders’ agents.
AI-written summary · about
chapters (15)
- 0:00Introduction to RSAC 2026
- 0:28Meet Roger Grimes and KnowBe4
- 1:09Human Risk Management explained
- 2:34AI agents and compliance risks
- 3:40Workforce trust management
- 5:04AI-powered cyber attacks
- 6:10Attacks against AI agents
- 7:55AI in cybersecurity defense
- 8:42AI infrastructure vulnerabilities
- 9:43AI-driven security training
- 11:30AI deep fake phishing demos
- 12:35Hyper-personalized training
- 13:41AI efficiency in security
- 14:24AI’s impact on security roles
- 16:11Final recommendations on AI
