For years, predictions about the future of commerce have often ended the same way: Credit cards eventually disappear.
Digital wallets would replace them. Cryptocurrencies would replace them. Buy now, pay later services would replace them. More recently, artificial intelligence has been cast as the next disruptor, with AI agents expected to handle everything from product discovery to checkout.
If an AI assistant can compare prices, find the best deal, and complete a purchase on your behalf, what role is left for a traditional payment network?
Visa thinks the answer is simple: a very big one.
At Visa Payments Forum 2026 this week, the company unveiled a slate of new AI, tokenization, and stablecoin initiatives designed to ensure that as AI transforms commerce, Visa cards remain the preferred way to pay.
In fact, the company is making a bet that the AI shopping boom could strengthen the role of credit cards, rather than weaken it.
AI is helping people shop—but it isn’t buying much yet
The vision of AI agents handling purchases for consumers has generated enormous excitement across the technology industry. But Visa says there’s a significant gap between consumers’ using AI to research products and actually allowing AI to spend money.
“I was literally in a Visa office overseas, a few months ago, and I asked for a show of hands. Tell me, everybody, if you’ve used AI to help you shop. Every single hand in the room goes up,” Jack Forestell, Visa’s chief product and strategy officer, tells Fast Company. “Tell me if you’ve used AI to actually check out. No hands go up. And this is a room full of payments nerds.”
Forestell said Visa identified the potential for AI-powered shopping more than a year ago as search tools became increasingly conversational and capable of helping consumers reach decisions faster.
“We really got focused on this whole notion of agentic commerce, or AI-assisted commerce, more than a year ago,” he said. “Our hypothesis was that ultimately that would end in agents actually not just helping with the discovery process but helping with the checkout process and post-transaction process itself.”
The problem isn’t that AI can’t help people shop. It’s that consumers aren’t yet convinced they should trust AI with their wallets.
The real barrier is trust
According to Forestell, technology is only part of the equation. Consumer trust remains one of the biggest obstacles preventing AI-powered commerce from becoming mainstream.
“If you were to have someone who you hardly knew go buy something for you, and you gave them access to your money to do it, like, how comfortable would you be with that value proposition?” he said. “I’d say you probably go with a no on that one.”
That hesitation helps explain why Visa sees an opportunity rather than a threat. The company isn’t trying to replace the card ecosystem for AI transactions. Instead, it’s building technology designed to make AI purchases feel as secure and protected as traditional card transactions.
“I think the consumers are going to want to understand that they are in control and they’re going to want to understand that they’re protected,” Forestell said.
Consumers already know that if something goes wrong with a purchase, they have recourse through their card issuer and the payment network. Visa believes those protections become even more important when software agents are making transactions.
“I need to be in control, even though I’ve actually granted some autonomy to this machine to actually execute the payment,” Forestell said. “And I want to make sure that if either the machine messes up or the merchant on the other side of that messes up, you still have my back.”
Why Visa is making credit cards smarter
Much of Visa’s latest announcement centers on a technology consumers rarely think about: payment tokens.
Today, many online transactions use tokens instead of card numbers. These digital credentials help protect payment information while allowing transactions to flow through existing card networks.
Visa is now adding more information to those tokens, including data about who initiated a transaction, where it originated, and whether an AI agent was involved. Forestell says that additional context is critical if AI agents are going to make purchases safely.
“It doesn’t necessarily tell you, oh, this was a purchase that was made using an AI agent, this is the identity of that AI agent, this is the trust score or the assurance score of that AI agent,” he said. “Those are the kinds of variables that we’re adding to tokenization.”
The goal is to help banks better understand what’s happening behind a transaction, improving fraud detection while reducing false declines. For consumers, that could mean fewer situations where legitimate purchases are mistakenly flagged while maintaining strong protections against fraud.
AI shopping still needs a payment system
Many discussions about AI commerce assume entirely new payment rails will emerge. Visa is betting otherwise.
The company announced a partnership with OpenAI that would allow AI agents to initiate Visa payments within user-defined permissions and controls. It is also launching an agentic directory, a registry of verified merchants and AI agents, along with tools that help merchants determine whether their websites are prepared for AI-driven shopping.
Taken together, the initiatives suggest Visa wants to become the trust layer underneath agentic commerce. The strategy makes sense, because AI agents still need a way to pay. Every transaction requires authorization. Every transaction needs fraud controls. Every transaction may require dispute resolution if something goes wrong. Those are areas where traditional payment networks already operate at massive scale.
The future may look familiar
Forestell believes some of the earliest AI shopping will involve routine purchases consumers don’t particularly enjoy making themselves.
“We’re seeing definitely some comfort” around AI handling “utilitarian, low- emotion, high-frequency, and low-value purchases,” he said.
Travel is another area where AI could gain traction. “We’ve had travel agents for 100-plus years for a good reason,” Forestell said. “Travel can be a very complicated process of planning and discovery and research.”
In both cases, however, the transaction still needs a trusted payment credential behind it. That’s why Visa doesn’t appear worried that AI will make cards irrelevant. Quite the opposite—while artificial intelligence may dramatically change how consumers shop, discover products, and make decisions, Visa is positioning itself so that the payment still flows through the same infrastructure consumers already know and trust.
The future of shopping may be powered by AI. But if Visa gets its way, the future of paying for it will still look a lot like a credit card.