
Microsoft has announced that its cloud-based development platforms, Azure AI Foundry and Copilot Studio, will now support Google’s recently introduced A2A (Agent-to-Agent) protocol, enabling seamless cross-cloud interactions between AI agents.
Concurrently, Microsoft will join the A2A protocol working group currently hosted on GitHub, contributing further to the development and deployment of applications based on this standard.
In addition to Microsoft, major industry players such as Accenture, Deloitte, SAP, Salesforce, and KPMG have already pledged their support for the A2A protocol, enabling their AI agents to interoperate with others—even when built using different programming languages, frameworks, or cloud infrastructures—through a unified communication standard.
Microsoft’s commitment to integrating A2A protocol support into Azure AI Foundry and Copilot Studio signifies a major step forward in facilitating interoperability across AI services within the Azure ecosystem and beyond, including services hosted on Google Cloud.
Moreover, Microsoft has previously integrated support for the Model Context Protocol (MCP)—an open architecture championed by AI company Anthropic—into Copilot Studio. This allows AI services to communicate and access data through standardized, interoperable requests.
By adopting both A2A and MCP protocols, Microsoft envisions a future where AI agents operate fluidly across diverse platforms and environments, enabling richer, more collaborative intelligence across the digital landscape.