Comment on Researchers jailbreak AI chatbots with ASCII art -- ArtPrompt bypasses safety measures to unlock malicious queries

<- View Parent
vamputer@infosec.pub ⁨3⁩ ⁨months⁩ ago

And then, in the case of it explaining how to counterfeit money, the AI gets so excited about solving the puzzle, it immediately disregards everything else and shouts the word in all-caps just like a real idiot would. It’s so lifelike…

source
Sort:hotnewtop