A made-up blog post claiming BBC technology journalist Thomas Germain was a "world champion tech journalist hot dog eater" managed to trick major AI tools like ChatGPT and Google Gemini.
Within just 24 hours, these platforms started repeating the false claim as fact, showing how easily AI can be misled by things it finds online.
Chatbots often omit source attribution
This slip-up highlights how much AI models rely on real-time web data and how they often don't show where their info comes from.
As Germain pointed out, "the chatbots often did not credit his blog as the source of the claim."
This left users with no easy way to check where the information came from.
To avoid getting fooled, it's smart to ask chatbots for sources and double-check facts with trusted sites.