Google Transfers AI Agent Protocol to Linux Foundation for Competitive Fairness

The Linux Foundation has recently initiated the Agent-2-Agent (A2A) project on GitHub, following the transfer of an open-source protocol developed by Google and various industry partners. This protocol is designed to facilitate communication between AI agents.

During the Open Source Summit North America held on June 23, the Linux Foundation announced that Google has officially handed over the A2A protocol. Concurrently, the organization has teamed up with major tech players, including Amazon Web Services (AWS), Cisco, Microsoft, Salesforce, SAP, and ServiceNow, to launch the A2A project hosted on GitHub.

The A2A protocol, developed by Google in collaboration with other industry leaders such as Atlassian, PayPal, Accenture, and McKinsey, aims to create a framework for AI agents to interconnect seamlessly. The Linux Foundation and its partners are committed to ensuring that A2A remains vendor-neutral, fostering fair competition and innovation in the AI space.

With support from over a hundred companies, including recent additions like AWS and Cisco, the A2A initiative seeks to encourage community participation for its ongoing development. The open-source nature of A2A complements Anthropic's Model Context Protocol (MCP) and enables AI agents to recognize one another, securely exchange information, and collaborate independently of their respective platforms, vendors, or frameworks.

Google had previously announced plans to launch a production-ready version of the A2A protocol within the current year, indicating a strong commitment to advancing interoperability among AI systems.