A2A joins the Agentic AI Foundation — horizontal agent interoperability is becoming infrastructure

The Agent2Agent protocol entered a new phase on 27 August 2026 when it was accepted as a Growth Stage project in the Agentic AI Foundation.

MCP gives an agent a standardized way to reach tools and data, while A2A gives independent agents a standardized way to discover one another, exchange capabilities and coordinate tasks.

For Outborn, A2A should complement rather than replace Agent Connect. Discovery must stay separate from authority, with authentication, authorization, rate limits, delegation context and evidence enforced by the control plane.

On 27 August 2026, A2A joined the Agentic AI Foundation as a Growth Stage project. It focuses on agent-to-agent discovery and task delegation, while MCP addresses tools and context. Combining both creates new identity and handoff questions: a downstream agent must not inherit broader authority than the initiating user.

Delegate a procurement request from an internal assistant to an external supplier agent. Verify identity, accepted task boundaries, callback authentication, cancellation and trace continuity.

Conceptual illustration; a real implementation must be tested separately.

Test handoff acknowledgement, authorized delegation, failure propagation and evidence linking the original request to final output.

Official reference: Primary publication. Avoid applying these conclusions outside the relevant legal, protocol or product version.