This chapter explores the learning mechanisms that enable Agentic Artificial Intelligence (AI) systems to adapt over time in dynamic environments. It focuses on Reinforcement Learning (RL) and multi-agent learning frameworks that support continuous policy refinement through interaction and feedback. The chapter examines challenges such as non-stationarity, coordination stability, and credit assignment in multi-agent settings. Learning is framed as an ongoing process operating across multiple temporal scales, enabling agents to balance short-term performance with long-term objectives. The integration of learning with language-based reasoning and coordination is discussed as a means of enhancing adaptability and resilience.