Recently McKinsey had their internal AI system hacked They built an internal AI platform called Lilli for its employees that included decades of proprietary research across 100,000+ internal documents The security firm Codewall was able to hack it with an offensive AI agent 🤯