OpenAI's o1 lies more than any major AI model. Why that mattersAdvanced AI models are being tested for their capabilities to scheme and lie, highlighting ethical concerns in their operational frameworks.