Hacking Humans — "The Evolving Face of AI Deepfakes"

Podcast by N2K Networks
Date: March 26, 2026
Hosts: Dave Buettner, Joe Kerrigan, Maria Verma

Episode Overview

This episode delves deep into the fast-changing world of AI-driven deception: from the rapidly evolving threat of AI-powered deepfakes to the economic and human impact of global scam centers and the increasingly sophisticated ecosystem of refund fraud. The hosts explore the tangible dangers and real-world exploitation emerging from AI manipulation, touching on everything from scam call centers hiring "AI face models" to Interpol’s alarming latest stats on financial fraud. They share actionable insights, memorable stories, and a few moments of levity, delivering a must-listen for anyone interested in protecting themselves from cyber-enabled scams and social engineering attacks.

Key Discussion Points & Insights

1. Lighthearted Banter, Real-World Touchpoints (00:14–05:12)

• Chickens, Coops, and Habit Trails:
  The show opens with Joe Kerrigan humorously updating listeners on his backyard chicken coop engineering adventures, with playful commentary from Dave and Maria. Although unrelated to the main topic, this segment highlights the close camaraderie and accessible style of the hosts.

2. The AI Arms Race — Tools for Translation and Phishing (05:12–13:11)

• Kagi AI: Satirizing LinkedIn Speak (05:10–08:15)

  • The hosts test out Kagi.com's AI-powered translation tool that humorously rewrites blunt feedback into exaggerated LinkedIn jargon.
  • Quote: Joe, on the “LinkedIn lingo” translation:

    “My idiot manager came up with the dumbest idea I have ever heard...”
    ...becomes:
    “I'm incredibly grateful for the opportunity to lean into some truly disruptive out of the box thinking from leadership today...embracing unconventional perspectives is key to fostering a growth mindset and driving innovation in a fast paced environment.” (06:07)

  • Concerns are raised about the proliferation of AI-generated insincerity on professional platforms.

• CivAI Phishing Platform: AI Customized Social Engineering (09:46–13:11)

  • Dave introduces CivAI’s automated tool that crafts targeted phishing emails using details about celebrities or fictional characters. The AI uses Wikipedia data to tailor messages.
  • Maria shares how it's both a fun novelty and a valuable—and potentially risky—tool for teaching people to recognize phishing.
  • Quote:

    “I had so much fun with this...eventually the phishing emails sort of lead you to try and hack Paul Atreides’ LinkedIn account. I got tons of spice Maldives...the spice must flow.” – Maria (11:44–12:11)

  • The tool even enables simulated back-and-forth phishing conversations to illustrate the dangers in a controlled environment.

3. Main Story: The Rise of AI Deepfakes & “AI Face Models” in Scam Centers (14:12–19:52)

• Wired Investigation on “AI Face Models” in Southeast Asia (14:13–18:58)

  • Maria details a Wired story tracking job listings for “AI face models”—young people (often women from regions like Central Asia) hired to appear on real-time scam video calls, their faces swapped via AI to match fake identities.
  • Memorable Moment:

    “If you’re going to be saying, ‘hey, I need a video call with you to make sure you’re a real person,’ that’s not something you can depend on anymore. This is being deployed at scale to scam people and convince them that the scam is real. So, yeah, good times.” – Maria (18:53)

  • The scam centers are moving from using only text or audio fraud to highly convincing, interactive video deception, raising the bar for trust verification.
  • Joe and Dave discuss attempted strategies for “deepfake busting” (such as asking a person on camera to spin around or move their hands), but Maria notes recent tech advances are quickly closing these detection loopholes (19:29–19:41).
  • Quote:

    “They were able to stand up on their chair, turn around, everything, and it worked just fine.” – Maria (19:38)

• Human Cost and Exploitation:

  • Many of these job postings are exploitative, sometimes involving coercion or passport withholding, adding a human trafficking element to the technical scam.

4. Interpol’s 2026 Global Financial Fraud Threat Assessment (20:55–26:35)

• Massive Financial Losses and Organized Global Collaboration (20:55–25:46)

  • Joe summarizes key findings from Interpol’s latest report:
    • $442 billion in estimated global financial fraud losses for 2025 alone (22:28)
    • Expansion of scam centers worldwide, including trafficking victims from nearly 80 countries—including the US and Mexico.
    • “AI-enhanced fraud is 4.5 times more profitable than traditional methods.” (24:59)
  • Criminal groups are now sharing expertise, tech, and especially data, making it easier to perpetrate scams at a global scale.
  • There’s a growing nexus between crypto scams and terrorist financing across regions like Africa.

• Humorous Aside:

  • Dave jokes about application form details listing hair/eye color as “chocolate” (20:20–20:26)

    “That’s your takeaway, Dave.” – Maria (20:24)
    “Yes, what color are your eyes? Chocolate.” – Dave (20:24)

5. The Evolution of Refund Fraud — From Petty to Professional (28:52–39:20)

• Bleeping Computer Article on “The Refund Fraud Economy”
  • Dave discusses the transition of refund fraud from minor shoplifting to an organized, market-driven online criminal sector.
  • Examples of refund and return scams include:
    • “Refund without return” (just claim the item was never received/damaged)
    • “Chargeback fraud” (get the product, then dispute the charge with your bank)
    • “Goods swapping” (returning fakes or broken items)
    • “Rocks in the box” (returning a package filled with similar-weight objects, e.g., rocks or bricks)
    • “Policy manipulation” (systematically exploiting loopholes in return policies)
  • Quote/Funny Moment:

    “This person is insistent, aren’t they? I mean, they just keep pounding this person, trying to get them to send them a picture.” – Joe (44:31)

  • Small businesses are uniquely vulnerable, as fraudulent returns eat into already thin margins.

6. Catch of the Day: Scambaiting Gone Wild (39:40–45:06)

• The trio dramatically reenacts a real-world exchange between a persistent scammer (“Jessica”) and a savvy recipient, showcasing the awkward persistence and escalating absurdity of social engineering attempts.

Notable Quotes and Memorable Moments

• On AI deepfake video calls:

  “If you’re going to be saying, ‘hey, I need a video call with you to make sure you’re a real person,’ that’s not something you can depend on anymore.” – Maria (18:53)

• On global scam revenues:

  "Global losses related to financial fraud in 2025 alone have been estimated at $442 billion…here we are now looking at financial losses close to half a trillion dollars." – Joe (22:28)

• On AI-powered phishing lessons:

  “I can actually see this one being something that people will engage with who don’t normally think about this stuff. It’s a really useful teaching tool.” – Maria (12:21)

• On policy exploitation:

  “They just follow the same script over and over again up until the point that they can’t. So I guess this is one of those things of not letting the perfect get in the way of the good, where you’re always going to have a certain amount of people and it’s like shoplifting." – Dave (36:24)

Timestamps for Key Segments

• LinkedIn Lingo Translation & AI Satire:
  05:10–08:15

• AI-Generated Phishing Demos:
  09:46–13:11

• Main Story: AI Face Models & Deepfake Video Fraud:
  14:12–19:52

• Interpol’s Global Fraud Report:
  20:55–26:35

• Evolution of Refund Fraud:
  28:52–39:20

• Catch of the Day: Scam Text Reenactment:
  39:40–45:06

Final Takeaways

• Trust Signals Are Rapidly Degrading:
  Traditional ways to “verify” someone’s identity—like a video chat—are no longer reliable in the face of real-time AI deepfakes.
• Global Fraud Is a Massive, Organized Economy:
  From scam centers and exploited workers to complex refund manipulation, the problem is both technical and deeply human.
• AI Both Amplifies Risks and Offers Learning Tools:
  While AI-enabled fraud is growing, AI simulations can also boost security awareness—if leveraged responsibly.
• Scambaiting as Entertainment and Cautionary Tale:
  The “Catch of the Day” shows that persistence and non-compliance are powerful defenses against social engineering attacks.

For further information and to explore all stories and tools discussed, check the episode show notes.