Skip to content

Four emerging protocols define how agents connect, coordinate, transact, and pay: MCP and A2A for connection and coordination, ACP and AP2 for commerce and payment.

Untangling Agentic AI Protocols

As agents become part of products and services, a set of shared protocols is beginning to take shape. Four in particular are worth knowing. They can be grouped into two categories: those that govern how agents connect, and those that govern how they transact.

Connecting and coordinating

MCP (Model Context Protocol): defines how an agent can use external tools and data. Introduced by Anthropic in 2024, MCP is now part of the Linux Foundation’s Agentic AI Foundation and has been adopted across products including ChatGPT, Gemini, Microsoft Copilot and VS Code.

A2A (Agent-to-Agent Protocol): launched by Google in April 2025 and contributed to the Linux Foundation in June 2025, A2A defines how agents discover capabilities, communicate and coordinate work.

Transacting and paying

ACP (Agentic Commerce Protocol): describes the structure of the checkout conversation between agent and merchant. OpenAI launched it with Stripe in ChatGPT’s 'Instant Checkout' in September 2025, initially for ChatGPT users in the United States buying from US Etsy sellers.

AP2 (Agent Payments Protocol): Google announced AP2 in September 2025 with payments and technology partners, then donated it to the FIDO Alliance in April 2026. It is best understood as a trust and authorisation layer for agent-led payments, not a new payment rail.

A simple way to think of them is this: MCP and A2A cover connection and coordination; ACP and AP2 cover commerce and payment. ACP is the commerce and checkout layer behind Instant Checkout; AP2 is a payment-authorisation layer designed to work with A2A and MCP. The boundary between commerce and payment standards is still settling, but they need not be competitors.