Just as a consumer might instruct an AI assistant to find and purchase groceries within budget and dietary preferences, a B2B buyer can direct an agent to source components exclusively from approved suppliers, remain within negotiated terms and select the most efficient payment rails while the secure payment transaction is processed in the background.
The potential for agentic commerce for B2B is enormous: without the need for extended processing for accounts payable and receivable, businesses can expect faster settlements, better working capital management, increased insights for large enterprises and inclusiveness for small businesses and creators who may not have had access to business management systems. By connecting verified credentials to merchant networks and financial institutions around the world, agentic commerce can easily scale across industries and geographies.
Key to all of this is payment credential readiness. Many traditional payment methods such as account and routing numbers, pay handles linked to wallets or file‑based bank transfers, are ill‑suited for autonomous agents. By contrast, card credentials and network tokens are already standardized, secure, and API‑ready, making them seemingly ideal for agent‑initiated flows.
Of course, the addition of a new participant in commerce, let alone a piece of AI software that can act on behalf of a human or a business, understandably triggers trust concerns. This article series will look to answer the core questions around trust and process that suppliers, issuers and others involved with agentic commerce likely have, from accounts payable to accounts receivable to the transaction itself.
The series looks to consider the involvement of agents in potential scenarios:
Before transactions
How do those involved in payments plan to verify agents? How do companies distinguish legitimate actors from malicious bots? How do participants ensure agents can access merchant systems without exposing them to security risks?
During transactions
How can everyone involved ensure agents are acting in line with the purchaser’s intent? How do companies prevent agentic “friendly fraud” wherein cardholders claim purchases are fraudulent and dispute them, kicking off the chargeback process, even though the purchases were likely originally approved? And how do companies make sure products and services are discoverable and accurately represented to agents?
After transactions
How do participants in payment processes hold agents accountable for outcomes? How do all players within the transaction process enable fast, fair dispute resolutions?
ABOUT THE AUTHOR
Mark Nelsen
Head of Product, Commercial & Money Movement Solutions, Visa
Mark oversees product development that helps people, businesses, and governments move money quickly and safely. He leads global product strategy for commercial credentials, digital money movement, and secure B2B payments. He has been with Visa for more than 20 years.