OpenAI Chain of Thought Controllability: Why AI Reasoning Models Struggle to Manipulate Their Thinking Process • Libertify