When a new 'agent commerce' standard shows up, should merchants rush to adopt it or just stick with the systems they already have?

This explores whether merchants preparing for AI shopping agents should adopt the new agentic-commerce protocols or keep using standards they already run. The corpus suggests the choice is mostly a false one. The new protocols that succeed tend to be thin layers built on old standards. Visa's Trusted Agent Protocol is a good example. It is pitched as new infrastructure, but it is built on HTTP Message Signatures, an existing web standard. Its new part is what the signature carries: the agent's intent, who the consumer is, and payment data Can cryptographic signatures help merchants identify trusted AI agents?. Coordination research outside commerce finds the same pattern. Standards get adopted when they wrap and connect existing protocols like MCP rather than asking everyone to rewrite their systems Should coordination protocols wrap existing systems or replace them?.

The 'which one' framing also breaks down because the new protocols aren't really rivals. Boboev describes ACP, AP2, UCP, Visa's TAP and Mastercard's Web Bot Auth as a ladder. Each one covers a different layer (identity, discovery, governance, checkout), and each absorbs a different level of fraud risk. Merchants stack them rather than pick one Do agentic commerce protocols compete or complement each other?. A more useful question for a merchant is how much agent risk they want to accept, and which layer handles it.

There is a real gap that existing standards don't fill. When Amazon blocked Meta's Muse shopping agent, it did so through its terms of service, because no agreed way existed for an agent to prove who it is and what it is allowed to do What identity standards do shopping agents actually need?. The law has a similar gap. EU consumer protections assume a human clicking through checkout screens, with required buttons and disclosures. Busch argues these rules can't be enforced when agents transact directly with each other through APIs. He proposes new protocol-level duties, such as a withdrawal API that lets a buyer cancel Can EU consumer law rules survive agent-to-agent API transactions?. So some new protocol work is unavoidable, both for identity and for compliance.

There is also a warning from engineering practice. Adding a protocol in between can make systems less reliable. One production team found that tool access through MCP caused unpredictable failures, and switching to direct function calls fixed them. A survey of 306 practitioners found that 85% build custom agents rather than use frameworks Why do protocol-based tool integrations fail in production workflows?. For checkout, where a wrong action costs money, merchants may want to use protocols at the edges (identity, discovery) and keep the transaction path itself direct and predictable.

The part you might not expect is that this is less a technical decision than a vote on market structure. O'Reilly argues that MCP turned a winner-take-all chatbot race into a market where many vendors can compete Do protocols reshape markets by enabling multi-vendor competition?. Waller expects agentic commerce to split into open and closed systems, the way payment networks split into open-loop and closed-loop Will agentic commerce platforms adopt open or closed models?. The UK's Competition and Markets Authority (CMA) treats data portability as core market infrastructure, without which consumers get locked in Should data portability be infrastructure for agent-driven markets?. A merchant that only adopts one platform's proprietary protocol is helping build a closed loop. A merchant that adopts open, layered standards keeps more agents able to reach it.