What Is the New Agent2Agent (A2A) Protocol and How Will It Shape AI Agent Interoperability?

Recently at the North America Open Source Summit, the Linux Foundation announced the creation of the Agent2Agent (A2A) project together with Amazon Web Services, Cisco, Google, Microsoft, Salesforce, SAP and ServiceNow.

The A2A protocol is an open standard for communication and collaboration between different AI agents, providing a common language that enables agents to discover each other's capabilities, exchange information securely, and coordinate complex tasks. Over 100 companies already support the protocol, with AWS and Cisco as the latest validators.

Hosted under the neutral governance of the Linux Foundation, the Agent2Agent project will maintain vendor‑independence and community‑driven development, accelerating adoption of the A2A specification, SDK and developer tools.

At AWS, we believe agentic AI is critical to virtually every customer experience. We welcome A2A joining the Linux Foundation and look forward to creating broader opportunities for developers building AI applications. We plan to support the community through contributions and access to the most extensive agent frameworks, protocols, and services. – Swami Sivasubramanian, Vice President, Agentic AI, AWS

We are committed to an open, interoperable agent internet and have joined the A2A project as a founding member because community‑driven development is the fastest path to widespread agent adoption. We will integrate A2A support directly into AGNTCY’s core open‑source components to build a truly cross‑vendor interoperability layer. – Vijoy Pandey, General Manager and Senior Vice President, Outshift, Cisco

Open standards are essential, but they are only part of the picture. Microsoft is dedicated to combining open interoperability with enterprise‑grade capabilities to shape the future of agent AI, helping enterprises responsibly deploy agents at scale. We welcome A2A’s transition to a neutral nonprofit project and look forward to collaborating on the next chapter of open agent development standards. – Yina Arenas, Vice President, Azure AI Foundry, Microsoft

Key Goals of the Agent2Agent Foundation

Establish open standards: Advance the A2A specification and adoption to become the primary industry standard for AI agent interoperability, ensuring universal compatibility.

Cultivate a vibrant ecosystem: Grow a diverse global community of developers, researchers, and companies to accelerate innovation and applications built on A2A.

Ensure neutral governance: Provide a fair competitive environment for all contributors and consumers under the Linux Foundation’s trusted open‑governance framework.

Accelerate secure innovation: Encourage the development of new applications and services that leverage safe, collaborative AI agent capabilities.

Linux Foundation Executive Director Jim Zemlin said the foundation is excited to host the A2A project, which will ensure long‑term neutrality, collaboration, and governance, ushering in a new era of productivity driven by agent‑to‑agent interactions.

Google AI will continue working with partners and others in the field to develop broader open standards that complement A2A in areas such as trusted agent identity, delegated permissions, governance policies, agent security, and reputation.

The A2A project invites all organizations and individuals passionate about the future of AI to join the community and contribute. For more information and to get involved, visit the A2A GitHub repository.

https://github.com/a2aproject