A2A (Agent-to-Agent)
- Definition
- A protocol that enables AI agents built by different vendors to discover, authenticate, and collaborate with each other. A2A standardizes how agents delegate sub-tasks, share context, and return results across organizational boundaries.
- Why it matters
- Agents in isolation hit a ceiling fast. Real enterprise workflows span CRM, ERP, HRIS, and dozens of other systems owned by different teams. If your agents cannot talk to each other through a shared protocol, you end up duct-taping point-to-point integrations that break every quarter. A2A is the network layer that makes multi-agent orchestration practical. Companies betting on single-vendor agent stacks will find themselves locked in, while those adopting open agent protocols preserve optionality. The winners here are the ones who treat agent interoperability as infrastructure, not an afterthought.
- In practice
- Google introduced A2A in April 2025 alongside partners like Salesforce, SAP, and Atlassian. The protocol defines an Agent Card (a JSON manifest describing capabilities), a task lifecycle (submitted, working, completed, failed), and streaming support. In a demo, a recruiting agent built on one platform delegated resume screening to a specialized HR agent built on another, with both exchanging structured messages over HTTP. Early adopters report cutting integration time from weeks to days when both sides support the protocol.
We cover products & deployment every week.
Get the 5 AI stories that matter — free, every Friday.
Related terms
Agent
An AI system that can autonomously plan, use tools, and execute multi-step tasks on behalf of a user. Agents are the next major product paradigm after chatbots, with every major lab shipping agent frameworks.
MCP (Model Context Protocol)
An open standard (created by Anthropic) that lets AI models connect to external tools, data sources, and services through a unified interface. MCP is becoming the USB-C of AI integrations.
Agentic orchestration
The architecture pattern of coordinating multiple AI agents to accomplish complex tasks, with a supervisor agent routing work, managing state, and combining results from specialized sub-agents.
Tool use
The ability of an AI model to invoke external tools, such as web search, code execution, or database queries, to augment its capabilities. Tool use transforms models from knowledge stores into action-taking agents.
Know the terms. Know the moves.
Get the 5 AI stories that matter every Friday — free.
Free forever. No spam.