The field of multi-agent systems and agentic AI is undergoing significant transformations, driven by the development of innovative architectures and protocols that enable more efficient, adaptive, and secure collaboration among agents. A key direction in this field is the shift towards proactive, generative, and anticipatory decision-making, allowing agents to model environment evolution, predict other agents' behaviors, and engage in strategic reasoning. This paradigm shift is expected to unlock unprecedented possibilities for distributed intelligence, moving beyond individual optimization towards emergent collective behaviors. Notable papers in this area include:

• Accelerating Drug Discovery Through Agentic AI, which introduces a novel AI framework for automating the drug discovery pipeline.
• GenAI-based Multi-Agent Reinforcement Learning, which advocates for a transformative paradigm shift from reactive to proactive multi-agent intelligence through generative AI-based reinforcement learning.
• Aime: Towards Fully-Autonomous Multi-Agent Framework, which introduces a novel multi-agent framework designed to overcome the limitations of the prevalent plan-and-execute framework.