Google's Agent2Agent (A2A) protocol has formally joined the Linux Foundation-directed Agentic AI Foundation (AAIF), the governance body announced on August 20, placing the open standard for agent-to-agent communication under the same neutral home as Anthropic's Model Context Protocol.
Two Open Standards, One Foundation
A2A defines how autonomous agents built on different frameworks discover each other, negotiate tasks, exchange identity credentials and maintain state across organizational boundaries. Anthropic's MCP, contributed as a founding AAIF project, handles the complementary problem of connecting agents to internal tools and data sources. With both under the AAIF umbrella, developers now have a single neutral steward for the two most-adopted open protocols in the agent stack, reducing the risk that either becomes a single-vendor moat.
AAIF Momentum
The AAIF has grown from 49 members to more than 250 in under a year and counts Platinum signatories including AWS, Anthropic, Block, Bloomberg, Cloudflare, Google, Microsoft and OpenAI. A2A itself passed 150 supporting organizations earlier this year and has landed inside Microsoft Azure AI Foundry, Google Cloud Vertex AI and Amazon Bedrock AgentCore Runtime, giving the protocol production reach across the three largest public clouds.
Why It Matters For Physical AI
Robots, humanoids and industrial cells increasingly ship as fleets of specialized agents rather than monolithic control stacks. Under a shared A2A/MCP layer, a warehouse humanoid can broker work with an inventory agent, a safety monitor and a shop-floor scheduler without proprietary glue code. That interoperability layer is now what NVIDIA's own physical-AI stack, Anthropic's forthcoming Model Hardware Standard and hyperscaler agent runtimes are all aligning on. Related coverage: Anthropic's Model Hardware Standard preview, NVIDIA Cosmos world foundation models, and Salesforce and Anthropic's Claudeforce partnership.
Reporting based on coverage from The Linux Foundation, Axios, Forbes and Yahoo Tech.