The Deepfake Episode: AI Avatars, CEO Clones & Likeness | Attempted Intelligence Ep. 4

knowbe4 logo
KnowBe4 | Securing Humans & Agents
Nov 4, 2026
KnowBe4

When AI-generated video floods your feed, how do you train an entire workforce on deepfakes without creating content that feels fake?

In Episode 4 of Attempted Intelligence, hosts Harlan Parrott (VP of AI Innovation at KnowBe4) and Matt Duren (SVP of Engineering for AI & Data) welcome the show’s first guest: Isida Drake, SVP of Security and Compliance eLearning at KnowBe4. Isida leads the global team delivering security training across 36+ languages and integrating AI directly into course development.

Together, they dig into the practical realities of using AI for enterprise content: why AI avatars get mixed reviews, who owns your digital likeness, what happens when employees see their own CEO deepfaked, and why effective security awareness isn't about memorizing single scams—it's about training the behavioral reflex to pause, verify, and question the request.

KEY TAKEAWAYS FROM EPISODE 4:

  • AI Content Realities: Where generative video and avatars still fall short, and how to keep digital training engaging.
  • Ownership & Likeness: Navigating the ethical, legal, and operational questions around personal digital clones and voice models.
  • Training the Behavior, Not the Event: Why teaching employees to pause and verify context beats trying to spot pixel-level deepfake flaws.
  • The Evolution of Work: Why the real risk isn't AI taking your job, but someone who uses AI more effectively taking your place.

CHAPTERS / TIMESTAMPS:

00:00 - Introduction to Attempted Intelligence

01:05 - Welcome to Episode 4: Meet Isida Drake

02:41 - Where AI in Content Creation Still Falls Short

04:10 - How to Avoid AI Training Video That Feels Fake

10:44 - Who Owns Your Digital Likeness in the AI Era?

14:16 - Using AI as a Springboard for Creative Teams

18:10 - Will Someone Using AI Better Than You Take Your Job?

21:52 - Inside KnowBe4's Content Creation Agent

25:15 - What Actually Keeps Learners Engaged with Training?

28:27 - Real-World Deception: Matt Gets Fooled by an AI Container Ship Video

31:14 - Security Coaching: Train the Behavior, Not the Event

37:15 - Should Creative Professionals Worry About Generative AI?

42:32 - Executive Impersonation: How Employees React to a Deepfaked CEO

FEATURED QUOTES:
💬 “Really, what you need to train on is the behavior. And the behavior is to pause and ask yourself questions of, like, is this my job role? Is this somebody I interact with often?” — Isida Drake
💬 “I got got by something innocuous. But it could’ve been a family member that sent me a video that I thought was something real.” — Matt Duren
💬 “The places where AI is being inserted, it’s not taking away something that a human would have normally done. It’s actually going and solving a problem that basically would have been done more poorly or not at all.” — Harlan Parrott

RESOURCES & LINKS:

#Deepfakes #AIAgents #AITraining #AttemptedIntelligence #KnowBe4 #Cybersecurity

Copyright © 2026 OpsMatters™. All rights reserved.