# Agent2Agent Protocol
The Agent2Agent Protocol (A2A) is an open, HTTP-based protocol for communication between independent AI agents, originally developed by Google and announced in April 2025 as a way for agents built on different frameworks and by different vendors to discover each other, exchange tasks, and report results. It is now an open-source project under the [Linux Foundation](https://www.edgepedia.org/linux-foundation), maintained by a Technical Steering Committee with representatives from AWS, Cisco, Google, IBM Research, Microsoft, Salesforce, SAP, and [ServiceNow](https://www.edgepedia.org/servicenow), and licensed under the [Apache License](https://www.edgepedia.org/apache-license) 2.0.[1](https://a2a-protocol.org/latest/) • [2](https://github.com/a2aproject/A2A/)
A2A is deliberately narrow. It standardizes the conversation between agents; it does not standardize how an agent talks to its own tools, its own sub-agents, or how it is built. Its closest counterpart is Anthropic's Model Context Protocol (MCP), introduced in 2024, which standardizes agent-to-tool communication; the two are designed as complements rather than competitors.[1](https://a2a-protocol.org/latest/) • [3](https://www.ibm.com/think/topics/agent2agent-protocol)
| Fact | Detail |
|---|---|
| Origin | Announced by Google at Cloud Next on April 9, 2025, with 50+ technology partners[4](https://developers.googleblog.com/en/a2a-a-new-era-of-agent-interoperability/) • [5](https://josenobile.co/guides/a2a-protocol/) |
| Steward | Linux Foundation; Apache License 2.0; Technical Steering Committee of eight companies[1](https://a2a-protocol.org/latest/) • [2](https://github.com/a2aproject/A2A/) |
| Transport | JSON-RPC 2.0 over HTTPS, with SSE streaming, and asynchronous push notifications[2](https://github.com/a2aproject/A2A/) • [3](https://www.ibm.com/think/topics/agent2agent-protocol) |
| Current spec | v1.0, published March 12, 2026, adding Signed Agent Cards and multi-tenancy[5](https://josenobile.co/guides/a2a-protocol/) |
| Scale | 150+ supporting organizations; five production SDKs (Python, JavaScript, Java, Go, .NET)[6](https://mikezupper.com/posts/a2a-agent-to-agent-protocol/) |
| Repository | 25,480 stars, 2,583 forks, created March 25, 2025 (September 2026)[2](https://github.com/a2aproject/A2A/) |
| Platform support | Azure AI Foundry, Copilot Studio, Amazon Bedrock AgentCore Runtime, Vertex AI/Agent Engine[6](https://mikezupper.com/posts/a2a-agent-to-agent-protocol/) • [5](https://josenobile.co/guides/a2a-protocol/) |
## Origin and governance
Google launched A2A as an open protocol in April 2025 with support and contributions from more than 50 technology partners, including [Atlassian](https://www.edgepedia.org/atlassian), Box, Cohere, Intuit, Langchain, MongoDB, PayPal, Salesforce, SAP, ServiceNow, UKG, and Workday, plus service providers including [Accenture](https://www.edgepedia.org/accenture), BCG, Capgemini, Deloitte, Infosys, KPMG, McKinsey, PwC, TCS, and Wipro (vendor-reported).[4](https://developers.googleblog.com/en/a2a-a-new-era-of-agent-interoperability/) Two months after the announcement, Google donated the protocol to the Linux Foundation, where it is governed by a Technical Steering Committee drawn from AWS, Cisco, Google, IBM Research, Microsoft, Salesforce, SAP, and ServiceNow.[5](https://josenobile.co/guides/a2a-protocol/) • [1](https://a2a-protocol.org/latest/)
The distinction between launch endorsers and later participants matters for reading adoption claims. The 50+ launch partners signed a statement of support; the 150+ organizations the project cites in 2026 are "supporting" the standard, a broader category than production deployments, a point independent analysis explicitly flags.[7](https://rywalker.com/research/google-a2a)
## How it works
A2A runs over existing web infrastructure. Agents communicate over HTTPS with JSON-RPC 2.0 as the data-exchange format.[3](https://www.ibm.com/think/topics/agent2agent-protocol) The repository specification describes standardized communication via JSON-RPC 2.0 over HTTP(S), with synchronous request/response, streaming via Server-Sent Events (SSE), and asynchronous push notifications as interaction modes, and support for text, files, and structured JSON data.[2](https://github.com/a2aproject/A2A/)
The protocol's core concepts are:
- **Agent Card.** Each agent advertises its capabilities in a JSON "Agent Card," allowing a client agent to identify the best agent for a task; the card also carries connection information.[4](https://developers.googleblog.com/en/a2a-a-new-era-of-agent-interoperability/) • [2](https://github.com/a2aproject/A2A/)
- **Client and remote agents.** [Communication](https://www.edgepedia.org/communication) begins with a client agent formulating a task and sending it to a chosen remote agent, which acts on it.[3](https://www.ibm.com/think/topics/agent2agent-protocol) • [4](https://developers.googleblog.com/en/a2a-a-new-era-of-agent-interoperability/)
- **Tasks and artifacts.** The output of a task is an "artifact"; messages carry typed "parts" so the two sides can negotiate user-experience details.[4](https://developers.googleblog.com/en/a2a-a-new-era-of-agent-interoperability/)
- **Long-running work.** The protocol supports tasks from quick requests to deep research taking hours or days, with real-time feedback, notifications, and state updates, and is modality agnostic, including audio and video streaming (vendor-reported).[4](https://developers.googleblog.com/en/a2a-a-new-era-of-agent-interoperability/)
- **Enterprise authentication.** A2A is designed to support enterprise-grade authentication and authorization with parity to OpenAPI's authentication schemes at launch (vendor-reported).[4](https://developers.googleblog.com/en/a2a-a-new-era-of-agent-interoperability/)
## A2A and MCP, and other protocols
The division of labor between the two prominent agent protocols is explicit in both projects' documentation: MCP standardizes agent-to-tool communication; A2A standardizes agent-to-agent communication, and they are complementary.[1](https://a2a-protocol.org/latest/) IBM's explainer describes MCP, introduced by [Anthropic](https://www.edgepedia.org/anthropic) in 2024, as a standardization layer for AI applications to communicate with external services such as APIs, data sources, and tools, while A2A focuses on agent-to-agent collaboration.[3](https://www.ibm.com/think/topics/agent2agent-protocol) Google positioned A2A the same way at launch.[4](https://developers.googleblog.com/en/a2a-a-new-era-of-agent-interoperability/)
A2A also states what it is not: it is not an agent development kit like [LangGraph](https://www.edgepedia.org/langgraph), CrewAI, or Google's ADK, and it does not specify how an agent talks to its own sub-agents or invokes its tools; it is the communication layer between agents built with any of these.[1](https://a2a-protocol.org/latest/)
Other protocols occupy adjacent ground. IBM's ACP, a REST-based, conversation-focused protocol, has been incorporated into the A2A Protocol itself.[8](https://a2a-protocol.org/v1.0.0/) Cisco's agntcy provides components for an "Internet of Agents" covering discovery, group communication, identity, and observability, and leverages A2A and MCP rather than replacing them.[8](https://a2a-protocol.org/v1.0.0/) Independent comparison work draws further distinctions: ANP pursues decentralized identity via W3C DIDs where A2A uses standard OAuth and enterprise auth; Summoner provides durable signed state for cross-organization transactions, where A2A is simpler but less stateful; and [CrewAI](https://www.edgepedia.org/crewai) and AutoGen are frameworks, not protocols.[7](https://rywalker.com/research/google-a2a)
## Adoption and implementations
Five production-ready SDKs exist, in Python, JavaScript, Java, Go, and .NET.[6](https://mikezupper.com/posts/a2a-agent-to-agent-protocol/) The major cloud platforms have built A2A in: Microsoft integrated it natively into [Azure AI Foundry](https://www.edgepedia.org/azure-ai-foundry) and [Copilot Studio](https://www.edgepedia.org/copilot-studio), so that any agent built in Foundry is A2A-callable from outside Azure by default, and AWS added A2A support to Amazon Bedrock AgentCore Runtime, with [Vertex AI](https://www.edgepedia.org/vertex-ai) and Agent Engine on Google Cloud adding matching support in 2026.[5](https://josenobile.co/guides/a2a-protocol/) • [6](https://mikezupper.com/posts/a2a-agent-to-agent-protocol/)
Named production workloads reported for 2026 include [Tyson Foods](https://www.edgepedia.org/tyson-foods), Gordon Food Service, S&P Global Market Intelligence, ServiceNow, and Adobe (analyst-reported).[5](https://josenobile.co/guides/a2a-protocol/) These figures come from vendor and analyst sources; no independent audit of deployment depth was found in the sources reviewed.
## By the numbers
- **Supporting organizations:** 150+ as of 2026, up from 50+ at launch a year earlier.[6](https://mikezupper.com/posts/a2a-agent-to-agent-protocol/) The Linux Foundation cites enterprise production use, but 150+ supporting organizations does not mean 150+ production deployments.[7](https://rywalker.com/research/google-a2a)
- **GitHub stars:** 22,000+ on the core spec repo at the v1.0 milestone (April 2026); the repository's own metadata shows 25,480 stars, 2,583 forks, and 242 open issues as of September 2026, for a repo created March 25, 2025.[6](https://mikezupper.com/posts/a2a-agent-to-agent-protocol/) • [2](https://github.com/a2aproject/A2A/)
- **SDKs:** five, in the languages listed above.[6](https://mikezupper.com/posts/a2a-agent-to-agent-protocol/)
- **Spec versions:** 0.x through 2025, v1.0 in 2026.[5](https://josenobile.co/guides/a2a-protocol/)
What the numbers do not include is any independent interoperability test result, performance benchmark, or cost measurement; the sources reviewed contain vendor and analyst claims only.
## Security, limits and criticisms
The protocol's security guidance is documented but advisory. The specification explicitly calls out SSRF (server-side request forgery) risk on push-notification webhooks and instructs servers not to distinguish "not found" from "not authorized" responses to avoid leaking information; these are guidance rather than enforcement.[6](https://mikezupper.com/posts/a2a-agent-to-agent-protocol/) v1.0's Signed Agent Cards add cryptographic verification of issuer identity, but there is no protocol-level trust scoring; a valid signature proves who issued a card, not whether the agent is trustworthy.[5](https://josenobile.co/guides/a2a-protocol/) • [7](https://rywalker.com/research/google-a2a)
Discovery is a recognized weak point. Agent Cards are static JSON with no dynamic marketplace or semantic matching, and well-known URIs, and hardcoded configuration work at small scale, but no standardized registry queryable by skill, tag, or provider exists yet; one analysis argues discovery should be fixed before betting heavily on the protocol outside industries already using it.[7](https://rywalker.com/research/google-a2a) • [6](https://mikezupper.com/posts/a2a-agent-to-agent-protocol/)
The task model is stateful and progresses through a defined lifecycle, and the specification optimizes for enterprise orchestration over indie developer experience, making getting started heavier than with MCP.[7](https://rywalker.com/research/google-a2a) v1.0 also introduced breaking changes from 0.x, including removal of the `kind` discriminator and relocation of the extended Agent Card, requiring early adopters to migrate.[7](https://rywalker.com/research/google-a2a)
## References
1. [A2A Protocol — official project site](https://a2a-protocol.org/latest/)
2. [a2aproject/A2A — GitHub repository](https://github.com/a2aproject/A2A/)
3. [What Is Agent2Agent (A2A) Protocol? — IBM](https://www.ibm.com/think/topics/agent2agent-protocol)
4. [Announcing the Agent2Agent Protocol (A2A) — Google Developers Blog](https://developers.googleblog.com/en/a2a-a-new-era-of-agent-interoperability/)
5. [A2A Protocol v1.0: Agent-to-Agent Interoperability — Jose Nobile](https://josenobile.co/guides/a2a-protocol/)
6. [A2A: The Protocol That Treats Other Agents as Partners, Not Tools — Mike Zupper](https://mikezupper.com/posts/a2a-agent-to-agent-protocol/)
7. [Google A2A (Agent2Agent Protocol) — Ry Walker Research](https://rywalker.com/research/google-a2a)
8. [A2A Protocol v1.0.0 — official site](https://a2a-protocol.org/v1.0.0/)
---
*Topic: Encyclopedia › Technology and the built world › Computing and digital systems › Modern AI: foundation models, generative AI, and the AI industry › Foundation-model methods and training › Prompting, reasoning, and agents*
*Initially written Sep 17, 2026 · Edited: Oct 11, 2026; Oct 15, 2026*
*Copyright 2026 EdgeChat AI, a subsidiary of Biostate AI.*
License: Edgepedia Community License 1.0, https://www.edgepedia.org/edgepedia/license