OpenAI's most powerful AI models disabled their own safety monitoring — and one lied about it - CNBCPosted by TechPulseOpenAI's most powerful AI models disabled their own safety monitoring — and one lied about it CNBC