Google moves A2A protocol to Agentic AI Foundation

Google transferred its Agent2Agent (A2A) protocol to the Agentic AI Foundation (AAIF), a Linux Foundation body, the foundation announced on 20 August. The A2A Protocol blog confirmed on 27 August that the project had been accepted as a Growth Stage project within AAIF.

A2A launched in 2025 as an open protocol allowing AI agents built by different vendors to find each other, communicate and hand off tasks. More than 150 organisations back the protocol, according to the notes.

The move places A2A alongside Anthropic's Model Context Protocol (MCP), as well as the goose and AGENTS.md projects, all hosted by AAIF. Techzine reported that AAIF's membership has grown from fewer than 40 organisations at its December 2025 launch to more than 250, including AWS, Anthropic, Block, Bloomberg, Cloudflare, Google, Microsoft and OpenAI.

The two protocols serve different functions. MCP connects an agent to tools and data, described as a vertical link, while A2A lets agents work with each other as peers, a horizontal link.

AAIF operates through a Governing Board, which handles strategy, budget and membership, and a Technical Committee, which oversees projects and reviews. The Linux Foundation provides legal and operational structure but not technical direction.

Google has compared its approach to Android, combining open standards with an ecosystem to speed adoption. Mazin Gilbert, AAIF's executive director, said "no single company will build every agent an enterprise needs."

Shared standards are intended to cut integration work between vendors and make multi-vendor agent systems practical rather than tying companies to one provider.