Agent2Agent protocol (A2A)
Also known as: A2A protocol, Agent2Agent, agent-to-agent protocol
The Agent2Agent protocol (A2A) is an open standard that lets AI agents built by different vendors communicate with each other, share tasks, and exchange results.
An agent publishes an 'agent card' describing what it can do, and other agents can send it work. While MCP connects an agent to tools and data, A2A connects agents to other agents, and in practice many systems use both.
Agent2Agent protocol (A2A), explained at three levels
At home
Imagine your travel-booking AI could ask a hotel's AI about room availability directly. A2A is a shared language that lets AI helpers from different companies talk to each other.
At work
A2A lets businesses combine AI agents from different vendors, such as a sales agent and a billing agent, so they can hand off work.
At home
Each agent has an 'agent card,' like a business card, listing its skills and how to contact it. Another agent reads the card and sends it a task, then gets updates and results back.
At work
A2A uses web standards such as HTTP and JSON-RPC, supports long-running tasks with status updates and streaming, and handles different content types like text, files, and forms. Agents stay 'opaque,' meaning they don't have to reveal their internal tools or reasoning.
At home
When agents talk to agents, it gets harder to see who decided what. In my opinion, that's why people should stay in the loop for anything important, like payments.
At work
Enterprise issues include authenticating agents across organizations, authorization and delegation of user permissions, audit trails across agent chains, liability for actions, and preventing one compromised agent from manipulating others. Standards maturity and vendor support should be checked before committing.
Real-world examples
At home
Fatima's AI assistant planned a dinner by checking a restaurant's booking agent for open tables, then confirmed with her before reserving.
At work
Marcus is a solutions architect at a Vancouver logistics firm. He piloted A2A so the company's customer-service agent could ask a separate shipment-tracking agent, built by another vendor, for delivery updates.
How it compares
| feature | Agent2Agent protocol (A2A) | Model Context Protocol (MCP) |
|---|---|---|
| Connects | Agents to other agents | An agent to tools and data |
| Originally from | Google, 2025 | Anthropic, 2024 |
| Governance | Linux Foundation | Agentic AI Foundation (Linux Foundation) |
| Typical use | Delegating tasks between agents | Calling tools and reading data |
MCP gives an agent hands. A2A gives agents colleagues.
For analysts
Multi-agent setups add complexity fast. I'd start with clear handoffs and audit logs, and only add agent-to-agent communication when one agent truly can't do the job alone.
History & origin
As companies began deploying AI agents from multiple vendors in 2024 and 2025, those agents had no common way to coordinate. Google announced the Agent2Agent protocol in April 2025, with support from more than 50 technology partners. In June 2025, Google transferred the protocol, its SDKs, and developer tools to the Linux Foundation as an open-source project. The Linux Foundation reported in 2026 that more than 150 organizations supported the standard, and official SDKs exist for Python, JavaScript, Java, Go, and other languages. Other proposals, such as IBM's Agent Communication Protocol, competed in the same space. In my view, A2A is promising, but most businesses should get one agent working reliably before worrying about agents talking to agents.