Linux Foundation takes over the A2A protocol

Background

A software agent is a program that uses a model to plan and carry out tasks on its own, often calling other software along the way. As such agents spread, they increasingly need to work with agents built by other teams, other companies and other platforms. A protocol is the shared set of rules that makes this possible: what a request looks like, how a reply is formatted, how a failure is reported. Without a common protocol, every pair of agents needs its own custom integration, and the work multiplies with each new system.

A2A is an open protocol created by Google for secure agent-to-agent communication and collaboration ^[1]. Google launched the protocol in April 2025, with growing support from more than 100 leading technology companies ^[1].

What happened

On 23 June 2025, at the Open Source Summit North America in Denver, the Linux Foundation announced the Agent2Agent project ^[1]. The foundation said the protocol lets autonomous agents discover one another, exchange information securely and collaborate across systems ^[1]. Discovery is a large part of the point: an agent has to be able to find a counterpart it did not know about in advance, check what that counterpart can do, and then exchange the information the task requires — across systems that were not built by the same vendor.

Key details

The announcement was framed around governance. Jim Zemlin, executive director of the Linux Foundation, said the move ensures long-term neutrality, collaboration and governance ^[1]. Under the foundation, A2A remains vendor neutral and keeps its focus on extensibility, security and usability ^[1]. Those two statements describe the same arrangement from different angles: a protocol that many companies depend on is easier to trust when no single one of them controls the rules, and the design priorities are written down as commitments rather than left to the founding company's product plans.

Rao Surapaneni, a vice president at Google Cloud, said the protocol establishes an open standard for communication across platforms and systems ^[1]. The emphasis on platforms and systems is what separates a protocol from a product feature: the value comes from agents on different infrastructure being able to talk to each other, not from agents within a single stack exchanging messages.

Why it matters

Placing a protocol with an independent foundation is a common way to signal that it will outlive the company that started it. Companies building on A2A can plan further ahead if the specification, the governance and the decision-making process do not depend on one vendor's priorities. Neutrality is also what allows a specification to be adopted by competitors, which is the only way a communication standard becomes useful: a protocol that only one company's agents speak solves nothing.

The project also puts the questions that matter for agent interoperability — how agents are discovered, how information is exchanged securely, how the protocol is extended — into a setting where multiple parties have a say ^[1]. For the wider field, the announcement marks a step in the shift of agent infrastructure from individual product launches toward shared, openly governed standards.