McKinsey's in-house AI, Lilli, got hacked
Another day, another AI horror.
Came across this story on X that explains how McKinsey's internal AI Lilli, trained on 100 years of proprietary data, got hacked and tons of sensitive data got stolen. All because of a public API that didn't even require authentication.
The full story is published here on The Register that touches the topic in even more detail.