From Browsing to Buying: How AI Agents Are Reshaping Payments

Welcome to Fintech Forward, where we break down the ideas, technologies, and signals shaping the future of money.

This week, one of the clearest signals yet came from Visa.

The global payments giant quietly revealed that it has already completed hundreds of real transactions powered by artificial intelligence as part of a pilot program launched earlier this year. That might sound modest at first—but in fintech terms, it’s a meaningful milestone. It suggests we are moving from talking about AI-driven commerce to actually using it.

From Recommendation Engines to Transaction Engines

For years, AI in commerce mostly lived upstream. It helped you search, compare, and discover. Recommendation engines nudged you toward products. Chatbots answered questions. But the final step—actually paying—was still firmly human-controlled.

That boundary is starting to blur.

Visa’s pilot focuses on what the industry is calling “agentic commerce”: systems where AI agents don’t just advise users, but act on their behalf. Instead of searching for concert tickets, comparing prices, entering card details, and clicking “buy,” consumers can increasingly delegate those steps to software.

Think less “AI assistant” and more “AI operator.”

Shoppers are adding to cart for the holidays

Over the next year, Roku predicts that 100% of the streaming audience will see ads. For growth marketers in 2026, CTV will remain an important “safe space” as AI creates widespread disruption in the search and social channels. Plus, easier access to self-serve CTV ad buying tools and targeting options will lead to a surge in locally-targeted streaming campaigns.

Read our guide to find out why growth marketers should make sure CTV is part of their 2026 media mix.

Learn more.

Why Payments Matter More Than Search

Search and discovery are relatively forgiving spaces for experimentation. If an AI makes a poor recommendation, the cost is low. Payments are different. They require trust, security, and precision. A failed or incorrect transaction isn’t just inconvenient—it’s expensive.

That’s why Visa’s announcement stands out. Payments networks are among the most risk-sensitive players in the tech ecosystem. If they are comfortable letting AI agents initiate transactions, it signals growing confidence in both the technology and the controls around it.

According to Visa’s leadership, early use cases are focused on repeatable, predictable purchases—the kinds of transactions where consumer intent is clear and risk is easier to manage. Subscriptions, routine shopping, or high-demand ticket purchases are natural starting points.

Visa Isn’t Alone—and That’s the Point

Visa’s experiment is part of a broader industry shift.

Across fintech and e-commerce, major players are racing to define how AI agents interact with money. Mastercard has been testing similar capabilities. Amazon has been experimenting with AI-assisted purchasing flows. Payment platforms and AI-native companies are partnering to make transactions conversational, automated, and increasingly invisible.

What’s striking is the speed of convergence. AI companies need payments. Payments companies need AI. Neither can fully shape the future of commerce alone.

Consumers Are More Ready Than You Think

One of the quiet assumptions holding back automation has always been consumer comfort. Would people really trust software to spend money on their behalf?

Early data suggests the answer is yes—at least in limited contexts. Surveys indicate that a significant portion of shoppers already use AI tools during the purchasing process, whether for product research, comparison, or planning.

What’s changing now is the handoff of authority. Instead of stopping at advice, consumers are beginning to allow AI systems to act, within defined boundaries. Spending limits, merchant rules, and approval thresholds act as guardrails, not roadblocks.

This mirrors earlier transitions in fintech: from cash to cards, from manual transfers to autopay, from passwords to biometrics. Each shift felt uncomfortable—until it didn’t.

The Strategic Implications for Fintech

For fintech companies, agentic commerce raises new strategic questions:

• Who controls the customer relationship when an AI agent intermediates the purchase?

• How are permissions, liability, and dispute resolution handled when software makes decisions?

• What becomes of branding and loyalty when consumers no longer browse interfaces?

Payments infrastructure will need to evolve to support identity verification, transaction intent, and real-time authorization at the agent level—not just the human level.

Visa’s plan to expand pilots into additional regions and deepen partnerships suggests this isn’t a side project. It’s foundational work for a future where transactions are initiated by software as often as by people.

Looking Ahead

We are still early. Hundreds of transactions are not millions, and pilots are not products. But direction matters more than scale at this stage—and the direction is clear.

AI is moving downstream, closer to money.

For consumers, that could mean less friction and more convenience. For businesses, it means rethinking how value is created and captured. And for fintech, it signals the next competitive frontier: not just enabling payments, but enabling autonomous payments.

The question is no longer whether AI will change how we shop. It’s how quickly we’re willing to let it buy.

Until next time,
Fintech Forward

Tracking the signals before they become the standard.