Security researchers created an algorithm that turns a malicious prompt into a set of hidden instructions that could send a user’s personal information to an attacker.
This Prompt Can Make an AI Chatbot Identify and Extract Personal Details From Your Chats
