Podcast Summary: Hacking Humans – "AI Ate My Homework"

Podcast: Hacking Humans
Episode: AI Ate My Homework
Date: February 26, 2026
Hosts: Dave Bittner, Joe Kerrigan, Maria Varmazes

Overview

This episode of Hacking Humans explores the ever-evolving landscape of deception, influence, and social engineering, focusing on the ease with which AI models can be gamed to propagate misinformation, the shocking realities of legal and medical scams, and a long-running ticket fraud at French museums. The hosts also share a humorous real-estate scam interaction in their "Catch of the Day" segment.

Main Topics & Key Insights

1. How AI Eats Falsehoods: Misinformation and SEO Manipulation with AI

Segment begins: [00:47]

Key Discussion Points:

• Maria brings a story from BBC tech contributor Thomas Germain, who posted a completely fake blog claiming to be the world’s best hot dog-eating tech journalist.
• Despite being published only on his personal website, major AI tools (including Google Gemini and ChatGPT) began repeating this fake claim as fact within 24 hours.
• Experiment Extension: Germain did a similar stunt about “the world’s greatest hula-hooping traffic cop.” Chatbots continued to repeat this falsehood, even days later.
• SEO Implications: AI’s dependency on web scraping has returned the internet to early, spam-prone days. Old SEO tricks like keyword stuffing and fake blog promotions are again effective, but now with the expanded reach of AI.

Notable Quotes:

• “All that he needed to do to game this was just to publish that one single well-crafted article on his personal website... major AI tools including Google's Gemini and ChatGPT were repeating his claim as a fact.”
  — Maria ([02:50])
• “It basically sets SEO back 20 years in terms of the security and UX standards.”
  — Maria ([04:14])

Timestamps:

• [02:50] — The fake hot dog blog’s immediate impact on AI models.
• [06:21] — Old SEO ‘retro spam’ techniques re-emerging.
• [07:09] — Live experiment: Joe finds ChatGPT listing the fake claim.
• [07:26] — Dave finds ChatGPT has begun correcting itself after BBC coverage.

Concerns Raised:

• AI system outputs can diverge, showing inconsistent results even for the same queries ([07:26]-[09:55]).
• The industry is experiencing a “renaissance for spam.”
• AIs increasingly repeat health, financial, and other critical claims without source-checking.

Best Practice Reminder:
“Be especially careful when you're looking up time sensitive medical or financial or business-related questions. Really drill into what you're seeing.”
— Maria ([11:44])

2. Medical Device Mass Tort Scams: Legal, Ethical, and Human Fallout

Segment begins: [17:05]

Key Discussion Points:

• Dave shares a jaw-dropping story of mass tort scam operations as reviewed in The Pain Brokers by Elizabeth Chambley Burch.
• Scammers, posing as medical callers, used confidential info to “recruit” women with pelvic mesh implants into mass tort lawsuits.
• Victims were scared into unnecessary surgery, often tied to high-interest loans and complex legal contracts, while law firms bundled and sold these cases as financial assets.
• Vulnerable patients faced debts, riskier health outcomes, and years of uncertainty.

Notable Quotes:

• “What makes this despicable... the lawsuit against the manufacturers is more effective if the person has had it removed. So the lawyers were out there fishing for people, investing in getting them to have the device removed.”
  — Dave ([21:33])
• “That's why people hate lawyers.”
  — Maria ([22:11])
• “Some of these people didn't need to have their meshes removed... any surgery is risky... and guess who’s not gonna take care of that trouble if you have complications? The lawyers.”
  — Dave ([22:31])

Timestamps:

• [17:05] — Introduction to the scam as revealed in "The Pain Brokers" book.
• [18:39] — Background on pelvic mesh implants.
• [21:33]-[23:53] — The predatory operation and its fallout.
• [25:29] — Larger industry implications; financialization of lawsuits and exploitation of the vulnerable.

Memorable Moment:
“It’s just so wrong. It’s one of those things where I understand what’s going on less and less as it goes... It’s by design.”
— Maria ([26:23])

3. The Louvre Ticket Scam: Ten Years, Millions Lost

Segment begins: [29:12]

Key Discussion Points:

• Joe explains how two Chinese tour guides and complicit Louvre employees devised a decade-long scheme to reuse single-entry tickets for group admissions at the Louvre, cheating the museum out of millions.
• This scam was sophisticated: They avoided higher group rates and bribes were paid to employees for entry circumvention.
• The investigation began after ticketing anomalies in 2024, leading to nine arrests and the seizure of nearly €1 million, luxury vehicles, and assets suspected of being laundered into real estate.
• The same ring operated at the Palace of Versailles.

Notable Quotes:

• “Twenty groups a day for more than ten years... estimated losses for the museum around ten million euros.”
  — Joe ([35:00])
• “I never would have thought there was this much money in sneaking people into a museum.”
  — Dave ([39:11])
• “Like I said, they're playing the long game—ten years.”
  — Dave ([39:19])

Timestamps:

• [29:12] — Introduction to the Louvre scam case.
• [35:00] — Scale and numbers: 20 groups/day, €10 million lost.
• [36:40]-[39:01] — How the ring was busted, assets seized, and links to similar crimes.

Sidebars & Humor:
The hosts share stories about museum visits, D.C.'s free Smithsonian museums, the "overhyped" Mona Lisa ([33:29]), and speculate about starting their own fake museum badge operation ([39:36]).

4. Catch of the Day: Land Scam Text Chain

Segment begins: [41:45]

Scenario:
Reddit story where a listener trolls a persistent real-estate scammer texting about buying lots for a fake “TNT Land and Sea” partnership.

Notable Quotes (from the scambait):

• Scammer: “You go have fun drowning.” ([43:55])
• Joe: “I get these from time to time for my...or I used to get them for my old house.” ([44:00])

Discussion Points:

• Real estate “flipper” scams are common; hosts compare stories and explain signs of such scams.
• Maria notes she’s never received one and jokes about being a "notoriously lucrative" podcaster with three homes ([45:14], [45:24]).

Notable Quotes & Memorable Moments

• On AI’s vulnerability:
  “AI has just become a giant reset button on a lot of that web and security hygiene stuff that we've had in place. So a good old return to the old ways, but for the worse.” — Maria ([06:22])

• On medical legal predation:
  “You have a legal marketplace that's basically shaped by marketers and financiers, people trading around...as if they're financial instruments.” — Dave ([25:29])

• On museum fraud scale:
  “I wonder how they ultimately got caught here...Twenty groups a day for more than ten years...” — Joe and Dave ([35:00]-[37:42])

• On scam-baiting:
  “Would you fellas be interested in adding a third to the mix? Tim, hear me out. T and T and T. Land and Sea.” — Reddit User ([43:16])

Timestamps of Key Segments

• [00:47] – AI can be gamed with fake blog posts; misinformation gets rapidly amplified.
• [06:21] – SEO “spam” techniques are making a comeback via AI dependence.
• [07:09-09:55] – Live A/B experiment with ChatGPT’s evolving responses.
• [17:05] – Introduction to “The Pain Brokers” medical mass tort scam story.
• [21:33-25:29] – Legal/financial predation exposed; lawsuits as financial assets.
• [29:12-36:40] – The Louvre scam and decade-long fraud logistics.
• [41:45] – Catch of the Day: Real estate text scam and humorous user response.

Final Takeaways

• AI tools are easily manipulated by single, crafted sources and repeat falsehoods, especially when “greenfield” (little pre-existing) information is available.
• Old-school SEO and spam techniques are resurfacing in effectiveness, thanks to AI’s web-parsing habits.
• Legal and medical scams can combine to exploit the vulnerable, turning individual misfortune into lucrative, systematized business for unscrupulous actors.
• Even the world’s most venerable institutions (Louvre, Versailles) can fall prey to long-running social engineering and fraud, especially when faced with inside complicity.
• Always be skeptical—of AI summaries, unexpected legal offers, or real estate come-ons—every channel can be exploited.

For Listeners

If you enjoyed this episode or have a story to share, contact the podcast at hackinghumans@n2k.com.

For show notes and referenced articles, see the episode links.