This week saw major strides in AI agent collaboration, with big tech companies working together to make different AI systems work seamlessly. Microsoft announced it will use Google’s A2A protocol to let its AI agents talk to others across platforms like Azure and Copilot Studio. This means agents built on Microsoft’s tools can now team up with external ones, even those from rivals like Google, to handle tasks like scheduling meetings or sending emails.

IBM also joined the push, launching tools like watsonx Orchestrate to help businesses build AI agents in minutes. These agents can connect to over 80 apps, helping workers in HR, finance, and IT automate tasks without coding skills. IBM stressed that orchestration and integration are key to making AI agents useful in real-world business setups.

In another partnership, IBM and AWS revealed deeper teamwork to boost agentic AI, aiming to let AI agents reason and act more independently across cloud systems. Meanwhile, events like Microsoft’s AI Agents Hackathon in San Francisco highlighted how startups are creating new tools for multi-agent collaboration, focusing on security and low-code systems.