Visa Inc. is moving deeper into agentic commerce, teaming up with OpenAI to support artificial intelligence (AI) agent-led payments.
The payments giant announced the collaboration this week at its Visa Payments Forum in San Francisco.
Under the new partnership, Visa said it will integrate its payment capabilities into “OpenAI experiences.” That gives developers and merchants a new framework to verify and process purchases that AI agents make.
The tie-up also gives OpenAI another path toward agentic commerce after its earlier push to bring agentic checkout into ChatGPT was scaled back.
“As AI agents become active participants in the economy, Visa’s focus is to ensure transactions are trusted, secure and seamless,” said Jack Forestell, Visa’s chief product and strategy officer, in a statement.
How agent‑led payments will work through Visa and OpenAI
The companies did not say when the new payment integration will fully launch, which merchants will participate or what the checkout experience will look like.
The move comes as retailers, payment networks and AI companies increasingly prepare for more shopping inside conversational platforms. While agentic commerce is still early, McKinsey & Company estimates it could account for up to $1 trillion in U.S. B2C retail sales by 2030. It said global volume could reach $3 trillion to $5 trillion.
In an agent‑led model, shoppers can hand off a task to an AI agent — such as finding a product or completing a purchase — and the agent carries out some or all of the steps.
Visa has been moving in that direction since early last year.
In April 2025, it launched Visa Intelligent Commerce, giving developers tools to connect AI agents to its payment network. By October, Visa and Mastercard had each rolled out separate frameworks to handle agent-led transactions.
Speaking at the Wolfe Research FinTech Forum in March, Forestell described agentic commerce as a major shift for the payments industry.
“You’ve got an agent in the middle,” he said. “The agent needs an identity. You need to secure that identity, you need to validate it, you need to collect more data in order to be able to ensure the security — all that stuff. And we’re working on that.”
Why agent-led payments need guardrails
The OpenAI partnership builds on that work. Rather than offering payment tools for AI agents in general, Visa said it’s now bringing that infrastructure directly to OpenAI’s platform.
In essence, card networks, banks and merchants need to know who approved the transaction, what the agent was allowed to buy and who is responsible if something goes wrong.
Through OpenAI’s platform, payments will use tokenized Visa credentials, which replace card details with secure digital codes. Payments will also use real-time authorization and fraud monitoring.
Visa said users would be able to set guardrails as well. This includes how much an OpenAI agent can spend, which types of merchants it can shop at and when a human user must approve the payment.
“By integrating with Visa Intelligent Commerce, we’re building the infrastructure for secure, transparent, and user-controlled agentic transactions, helping people do more with AI agents while maintaining confidence that payments are being handled safely and securely,” Marco Mahrus, OpenAI’s head of partnerships and commerce, said in a statement.
OpenAI gets another path into commerce
OpenAI has also been preparing for an AI commerce future through its Agentic Commerce Protocol, which it co‑developed with payments vendor Stripe. The open technical standard is designed to support AI‑enabled transactions.
In September, OpenAI debuted Instant Checkout in ChatGPT. Powered by the protocol, the feature let users complete some purchases from participating merchants without leaving the chat interface. That rollout started with Etsy Inc. and remained limited to participating sellers.
By March, OpenAI said it was shifting those plans, moving Instant Checkout into individual apps “to better meet merchants and users where they are.”
By contrast, the Visa partnership could reach further as it focuses on the broader payment infrastructure rather than a single checkout flow.
“Commerce is going to happen in many more places and in many more ways than it does today,” Mahrus said in Visa’s announcement. Agents, he added, will play a growing role in helping people complete tasks involving money, including “more complex transactions.”
Visa adds more agentic commerce tools
Visa previously said it’s aligning its own agent framework, Trusted Agent Protocol, with OpenAI’s protocol.
Under the new partnership, the companies also plan to explore enterprise uses, including developer-centric experiences powered by Codex, OpenAI’s coding agent. Those efforts could also include “more automated and conversational workflows,” Visa said.
Alongside the OpenAI collaboration, Visa announced several more tools tied to its agentic commerce plans:
- Agent Score. Evaluates a merchant’s site for agentic commerce “readiness.”
- Agentic Directory. A registry of Visa-verified agents and merchants.
- Large Transaction Model. An AI model trained on billions of transactions to improve fraud detection and reduce false declines.
Mastercard also expands agentic commerce tools
Also this week, Mastercard introduced its own agentic AI service, with a focus on agent-to-agent transactions.
Called Agent Pay for Machines, or AP4M, the service is designed for a future where businesses offer services that AI agents can buy and use. Mastercard said those agents could eventually transact with one another continuously, including through chains of smaller payments or microtransactions.
Those payments could happen faster and in smaller amounts than most do today. Mastercard said AP4M will allow machine-to-machine transactions to be “permissioned, orchestrated and settled” across its global network.
“Agent Pay for Machines will create the conditions for a super-bloom of AI business models,” said Jorn Lambert, Mastercard’s chief product officer.
The service builds on Mastercard’s Agent Pay program, which it rolled out in 2025. Mastercard said it is working with partners — including Checkout.com, Cloudflare, Coinbase, Ripple and Stripe — to test use cases and establish common rules.
Do you rank in our databases?
Submit your data and we’ll see where you fit in our next ranking update.
Sign up
Stay on top of the latest developments in the online retail industry. Sign up for a complimentary subscription to Digital Commerce 360 Retail News. Follow us on LinkedIn, X (formerly Twitter), Facebook and YouTube. Be the first to know when Digital Commerce 360 publishes news content.
Favorite