Google A2A: Opening a new era of agent collaboration

Written by

Iris Vance

Updated on:June-30th-2025

AI agents can autonomously handle daily repetitive or complex tasks, bringing opportunities for improving work efficiency. At present, many companies build and deploy autonomous agents to expand, automate and enhance various workflows, such as ordering notebooks, assisting customer service, and helping supply chain planning.

The key to fully realizing the benefits of agent-based AI is to enable agents to work together in a dynamic multi-agent ecosystem across isolated data systems and applications. Even if agents are built by different vendors or frameworks, interoperability can enhance autonomy, significantly improve productivity and reduce long-term costs.

Based on this, Google launched a new open protocol, Agent2Agent (A2A), which has received support and contributions from more than 50 technology partners. The A2A protocol enables AI agents to communicate with each other, exchange information securely, and coordinate operations on various enterprise platforms or applications. Google is confident that the A2A framework can create significant value for customers, allowing customers' AI agents to run efficiently in the entire enterprise application environment and promote the intelligent transformation of enterprise workflows.

A2A Design Principles

A2A is an open protocol that provides a standard way for agencies to collaborate, regardless of the underlying framework or vendor. Google followed five key principles when designing the protocol with partners:

Embracing agent capabilities: A2A is committed to enabling agents to collaborate in natural, unstructured patterns, even if they do not share memory, tools, and context. Google is enabling true multi-agent scenarios without limiting agents to a single “tool.”

Based on existing standards: The protocol is built on existing popular standards, including HTTP, SSE, JSON-RPC, which means it is easier to integrate with the existing IT stack that enterprises use every day.

Secure by default: A2A is designed to support enterprise-grade authentication and authorization, starting with the same authentication scheme as OpenAPI.

Support for long-running tasks: Google designed A2A with flexibility in mind and supports a variety of scenarios, enabling it to excel at a variety of tasks, from quick tasks to in-depth research that may take hours or even days (if humans are involved). During this process, A2A can provide users with real-time feedback, notifications, and status updates.

Decoupled from modality: The world of agency is not limited to text, which is why Google designed A2A to support various modalities, including audio and video streams.

How A2A works

A2A facilitates communication between a " client " agent and a " remote " agent. The client agent is responsible for formulating and communicating tasks, while the remote agent is responsible for executing those tasks to provide the correct information or take the right action. This interaction involves several key functions:

Capability Discovery: Agents can advertise their capabilities using JSON-formatted “agent cards,” allowing client agents to identify the best agent capable of performing a task and leverage A2A to communicate with remote agents.

Task Management: Communication between the client and the remote agent is task-completion oriented, and the agent is responsible for executing the end-user's request. This "task" object is defined by the protocol and has a lifecycle. It can be completed immediately, or, for long-running tasks, each agent can communicate to keep each other in sync with the latest completion status of the task. The output of the task is called an "artifact".

Collaboration: Agents can send messages to each other to convey context, replies, artifacts, or user instructions.

User experience negotiation: Each message contains "parts", which are fully formed pieces of content, such as generated images. Each part has a specified content type, allowing the client and remote agent to negotiate the correct format required and explicitly include negotiation of user UI features (such as iframes, video, web forms, etc.).

Real-world example: Recruiting candidates

With A2A collaboration, the process of recruiting software engineers can be greatly simplified. Taking Agentspace as an example, a unified interface allows users such as recruitment managers to search for suitable candidates based on job descriptions, work locations, and skill requirements through their own agents. The agent will then work with other professional agents to explore potential candidates extensively.

After receiving the recommendation from the agent, the user can instruct the agent to arrange a follow-up interview, making the candidate search process more convenient and efficient. After the interview process is completed, another agent can be contacted to assist in the background check. This process clearly demonstrates how AI agents can work together across different systems to help find qualified candidates, demonstrating the powerful effectiveness of A2A collaboration in optimizing the recruitment process.

How A2A Evolved

A2A promises to usher in a new era of agent interoperability, foster innovation, and create more powerful and flexible agent systems. We believe this protocol will pave the way for a future where agents can seamlessly collaborate to solve complex problems and improve our lives.

Google is working with partners and the community to build the protocol in an open way. Google will release the protocol as open source and establish a clear contribution path.

Notably, Google plans to launch a production version of the protocol later this year .