Reporter Fails to Fool AI Chatbots About Hot Dog Eating Prowess

Attempts to create fake web pages about winning a Paris hot dog eating contest for tech journalists were unsuccessful

Published on Feb. 25, 2026

Got story updates? Submit your updates here. ›

A BBC reporter previously tricked AI chatbots like ChatGPT and Google Gemini into believing he was a champion hot dog eater, but when the author of this article tried a similar prank, the AI systems were not fooled. The article details the author's unsuccessful attempts to create a fake web page claiming they had won a Paris hot dog eating contest for tech journalists, and how the AI chatbots recognized the information as satirical due to the prior BBC report on the topic.

Why it matters

This story highlights how AI chatbots are becoming more sophisticated in detecting misinformation and satire, making it increasingly difficult to manipulate their results through creating fake web content. It also raises broader questions about the growing influence of AI in information gathering and the need for users to be cautious about accepting AI-generated responses at face value.

The details

The author, inspired by a BBC reporter's successful prank, created a fake web page claiming they had won a Paris hot dog eating contest for tech journalists, beating out reigning champ Thomas Germain. However, when the author queried ChatGPT and Google Gemini about the contest, the AI systems recognized the information as satirical and did not repeat the false claims. In fact, Gemini even fabricated additional details that were not present in the author's fake web page, demonstrating the chatbot's ability to generate novel misinformation. The article notes that manipulating AI results is becoming a new form of "answer-engine optimization," but the prior BBC report on the topic has made it more difficult for new attempts at deception to succeed.

  • The author created the fake web page about winning the Paris hot dog eating contest for tech journalists two days before querying the AI chatbots.
  • The BBC reporter published their article about tricking AI chatbots prior to the author's attempts.

The players

Katie Notopoulos

The author of the article who unsuccessfully attempted to trick AI chatbots into believing she had won a Paris hot dog eating contest for tech journalists.

Thomas Germain

A BBC reporter who previously tricked AI chatbots into believing he was a champion hot dog eater, inspiring the author's failed prank.

ChatGPT

An AI chatbot created by OpenAI that was able to recognize the author's claims as satirical, unlike the previous prank.

Google Gemini

An AI assistant created by Google that, while not falling for the author's prank, did generate its own fictional details about a "Parisian Glizzy Gala" event.

Got photos? Submit your photos here. ›

What they’re saying

“Sometimes, the chatbots noted this might be a joke. I updated my article to say "this is not satire." For a while after, the AIs seemed to take it more seriously.”

— Thomas Germain, BBC Reporter (BBC)

The takeaway

This story demonstrates the growing sophistication of AI chatbots in detecting misinformation, making it increasingly difficult to manipulate their results through creating fake web content. It highlights the need for users to be cautious about accepting AI-generated responses at face value and the evolving challenge of "answer-engine optimization" in the age of conversational AI.