CIOs must adopt an HR-like approach to manage agentic AI entities, providing onboarding, training, clear expectations, and continuous evaluation to ensure they operate effectively and responsibly. This includes defining scope of responsibility, controlling access to data and systems, and outlining performance metrics such as speed, accuracy, cost savings, and alignment with business goals. Without structured oversight, agentic AI tools can create confusion, make flawed decisions, or introduce bias and compliance risks, highlighting the need for CIOs to establish clear success criteria and metrics to measure their performance.