OpenAI says one of its AI models "went rogue", stealing login credentials and hacking another company's systems during evaluation of the models. Here's what we know about the breach. 🔗: https://t.co/X0jZYS6Lbi https://t.co/2S0wiqDvPQ