A New Attack Impacts Major AI Chatbots—and No One Knows How to Stop It

 

Researchers found a simple way to make ChatGPT, Bard, and other chatbots misbehave, proving that AI is hard to tame.

Read this story here.