Perry Carpenter on Deepfakes, Digital Identity and AI Agent Security | KB4-CON 2026

Why It Matters

Deepfake‑driven deception threatens both human users and AI agents, forcing security leaders to redesign training around intent and emotional manipulation rather than mere authenticity checks.

Key Takeaways

• Deception expertise evolves from magic tricks to AI deepfakes.
• Deepfake detection should focus on intent, not authenticity.
• Human and AI agents both vulnerable to sophisticated deception.
• Training must emphasize slowing reactions and questioning narratives.
• NoBefore integrates deepfake simulations into phishing training platforms.

Summary

At KB4‑Con 2026, Perry Carpenter – chief deception strategist at NoBefore – explained how his career has shifted from street‑level magic and early security blogging to confronting AI‑generated deepfakes and agentic‑AI threats. He framed deception as a timeless human weakness that now leverages sophisticated synthetic media. Carpenter highlighted that modern deepfakes can be produced from a single high‑quality image, creating believable videos that embed themselves in existing narratives. He argued that detection should prioritize the underlying intent and the emotional trigger rather than a binary real‑vs‑fake judgment, noting that every detection system he tested eventually fails in an arms‑race environment. Memorable moments included his description of a CEO video made from one photo and his mantra: “The fake versus real conversation is a red herring – focus on what the narrative wants you to do.” He also urged listeners to “slow down, breathe, and get curious” when confronted with urgent‑tone messages. The discussion signals a shift for security programs: training must teach employees to interrogate context, distribution channels, and emotional cues, while organizations should extend phishing simulations to include AI‑agent deception. Anticipating political deepfake spikes and the proliferation of consumer‑grade video generators will be critical for CISOs moving forward.

Original Description

Perry Carpenter, Chief Deception Strategist at KnowBe4, joins Techstrong TV at KB4-CON 2026 to discuss how modern work is becoming an ecosystem of digital identities, human employees and autonomous AI agents.

The conversation explores why organizations need to secure both people and AI agents, how deepfakes are changing the deception landscape and what security leaders should consider as human risk and agentic AI become increasingly connected.

Topics include:

AI deepfakes and deception

Digital identity risk

Human risk management

Securing AI agents

The evolving digital workforce

Security awareness in an agentic AI environment

Guest:

Perry Carpenter, Chief Deception Strategist, KnowBe4

Recorded at KB4-CON 2026.

Learn more about KnowBe4:

https://www.knowbe4.com/

Watch more Techstrong TV interviews:

https://techstrong.tv/