Holiday shopping has already moved from stores to websites to apps. The next shift may be from shoppers themselves to AI agents.
Agentic commerce is moving AI beyond recommending what to buy. Increasingly, the vision is for an AI agent to search for products, compare options, build a cart, and eventually complete a transaction on a consumer’s behalf.
That could make holiday shopping dramatically easier. Tell an agent you need five gifts under $50, specify your preferences, and let it do the searching.
But there is one big complication: money.
When AI Goes From Recommending to Buying
The difference between an AI assistant and an AI shopping agent is action.
Once an agent can initiate a payment, the financial system has to answer questions it was not necessarily designed for. Did the consumer actually authorize this purchase? Is the agent legitimate? Did it follow the shopper’s instructions? Who is responsible if something goes wrong?
Those questions are becoming increasingly urgent. In September 2026, major banks including Bank of America, Capital One, ING, NatWest, Commonwealth Bank of Australia, and ASB Bank raised concerns about the potential for AI shopping agents to create new risks around scams, fraud, data privacy, and consumer protection. (Source)
The concern is not necessarily that agentic commerce should not happen. It is that the infrastructure surrounding it needs to catch up.
Identity Is Becoming Part of the Payment
Traditional digital commerce generally asks whether the person making a transaction is legitimate.
Agentic commerce adds another layer: Is the agent legitimate too?
Payments networks are already building infrastructure around that question. Visa’s Trusted Agent Protocol uses cryptographic signatures to help merchants distinguish approved commerce agents from malicious bots and verify information about the consumer and payment behind a transaction. (Source)
Mastercard is similarly developing systems designed to establish “verifiable intent,” giving merchants a way to confirm what a consumer actually authorized an AI agent to do. (Source)
That means identity, authorization, and payments could become increasingly intertwined.
Trust Could Be the Biggest Barrier
Consumers may like the convenience of AI shopping without necessarily being ready to hand over their wallets.
A 2026 Visa survey found that just 23% of U.S. consumers trusted generative AI to handle payment transactions on their behalf. When respondents were told Visa would secure the transaction, trust increased to 61%. (Source)
That gap says a lot about where FinTech fits into the agentic commerce story.
The winning experience cannot simply be autonomous. It has to make authorization clear, protect sensitive payment information, preserve consumer control, and provide a path to recourse when something goes wrong.
A New Holiday Checkout Is Coming
Holiday shopping is an ideal testing ground for this new model. Consumers make more purchases, shop across more merchants, search for deals, and often have specific budgets and deadlines.
AI agents could make that process significantly easier.
But the more responsibility consumers hand to AI, the more important the infrastructure underneath becomes.
The next evolution of checkout may not begin with a consumer clicking “buy.” It may begin with an AI agent asking permission to do it for them.
And for FinTech, making that transaction trustworthy could be just as important as making it possible.
