OpenAI and Anthropic have huge teams working on this stuff. You’ve got six questions and your cybersecurity instincts.

Could you keep a rogue AI contained?