ROAS Suite

Visa Unveils Research on the AI-Driven Commerce Era

By Charles Ryder

I’ve been tracking AI’s role in commerce for some time, and Visa’s latest research makes one thing clear: AI is moving beyond a support role. We’re entering a stage where AI can act as the customer.

That’s the core idea behind Visa’s new Business-to-AI, or B2AI, research, released alongside its investor announcement, “Visa Defines the Next Era of Commerce: When AI Becomes the Customer.” For payments, e-commerce, retail, and digital marketing teams, this is more than another trend report. It signals that commerce infrastructure is being rebuilt for a world where autonomous agents search, compare, negotiate, and purchase on behalf of people and businesses.

Visual representation of AI agents interacting with commerce, symbolizing Visa's B2AI research and the future of AI as a customer. From AI Assistant to AI Buyer

The major shift is autonomy.

For years, AI has helped consumers discover products, summarize reviews, and recommend alternatives. Visa is pushing that idea further: what happens when AI doesn’t just advise the customer, but actually becomes the buyer?

That question has been building for months. In 2025, Visa launched Visa Intelligent Commerce, a platform built to let AI agents make secure purchases using Visa credentials. The company followed with whitepapers, regional expansion, and live pilots across multiple countries. By late 2025, Visa had already completed hundreds of real AI-agent transactions without human intervention, working with partners including Banco Santander, PayOS, and BeyondStyle.

This latest B2AI research feels like the next logical step. It goes beyond showcasing the technology. It maps the market opportunity and highlights the readiness gap between what businesses want and what consumers are prepared to trust.

The Research Signals a Market in Transition

What stands out most is how far adoption has already progressed.

According to Visa’s research, 77% of businesses are already using or piloting AI, and 71% expect to optimize products or services for AI agents. Even more telling, 53% are open to AI-to-AI negotiation. Businesses are starting to picture a future where software doesn’t just support commerce behind the scenes, but actively takes part in it.

On the consumer side, the picture is more mixed. People are generally comfortable using AI for price comparison and shopping assistance, and a meaningful share already does. But confidence drops when spending authority comes into play. Most consumers still want approval rights, spending limits, and the ability to override decisions. Only a small minority are comfortable with unrestricted AI spending.

That gap matters. Visa appears to understand that adoption won’t depend on capability alone. It will depend on trust architecture.

Trust Is Becoming Commerce Infrastructure

This is where Visa’s position looks especially strong.

Frank Cooper III, Visa’s CMO, has emphasized that visibility, control, and the ability to step in are not optional in AI commerce. They are prerequisites. That’s exactly the point. The real obstacle to agentic commerce isn’t whether AI can make purchases. It’s whether people trust the system to operate within clear boundaries.

That explains why Visa has focused so heavily on the rails around the transaction, not just the transaction itself. Through Visa Intelligent Commerce, the company has been building APIs for authorization, tokenization, and payment instructions that help keep AI behavior aligned with user intent. Its newly announced AI-powered dispute resolution tools extend that same logic, giving merchants, acquirers, and cardholders ways to handle the messy edge cases that inevitably come with autonomous transactions.

Put simply, Visa isn’t just betting on AI-powered checkout. It’s positioning itself as the trust layer for AI-to-AI commerce.

Why This Changes the Sales Funnel

If AI agents start shopping and buying at scale, the traditional sales funnel changes fast.

A human shopper can be influenced by branding, layout, emotion, urgency tactics, and visual design. An AI agent works differently. It needs structured product data, transparent pricing, clear policies, trusted credentials, and machine-readable signals. In that environment, the brands that win may not be the ones with the flashiest storefronts, but the ones best optimized for machine decision-making.

That has major implications for retailers and marketers.

Product pages may need to serve two audiences at once: humans and autonomous agents. Checkout may become less about persuasion and more about permissions, verification, and negotiation. SEO may also broaden, where discoverability depends not just on search engines, but on whether AI agents can accurately interpret and trust your offer.

For brands, this is a structural change, not a cosmetic one.

Infographic illustrating key findings from Visa's research on AI adoption in commerce, showing business and consumer readiness for AI agents. Visa’s Timing Looks Deliberate

Visa’s April 2026 B2AI report didn’t arrive in isolation. It followed a year of deliberate moves.

The company introduced the infrastructure in 2025, expanded into Asia Pacific, validated the concept through international pilots, and then supported it all with consumer and business research. That sequence matters because it shows Visa isn’t theorizing about agentic commerce from the sidelines. It is building, testing, and educating the market at the same time.

Competition is heating up as well. Mastercard is developing its own trust-centered frameworks, while enterprise AI companies and payment innovators are all trying to define the future stack. Still, Visa has one clear advantage: scale paired with credibility. In a market where trust will shape adoption, incumbents with global payment networks start ahead.

What Businesses Should Take Away Now

If I were advising brands today, I’d say this: prepare for AI as a buyer, not just as a tool.

That means asking practical questions:

  • Is your product data structured for machine interpretation?
  • Can your pricing, policies, and inventory be consumed by autonomous systems?
  • Do your commerce workflows support secure approvals and spending controls?
  • Are you ready for disputes initiated by or involving AI agents?
  • Can your acquisition strategy adapt if more purchase journeys start outside the traditional website funnel?

Companies that wait for AI commerce to become mainstream may end up redesigning under pressure. The ones that prepare early will be in a much stronger position to capture demand as this model scales.

FAQ

What is Visa’s B2AI research about?
Visa’s Business-to-AI research examines how commerce changes when AI systems act not just as assistants, but as autonomous buyers that can search, compare, negotiate, and complete purchases.

What is agentic commerce?
Agentic commerce refers to a model where AI agents carry out shopping and purchasing tasks on behalf of consumers or businesses, often with varying levels of autonomy.

Why is trust so important in AI-driven commerce?
Consumers may accept AI recommendations, but giving AI spending authority is a much bigger leap. Trust depends on clear controls, visibility, approvals, and reliable dispute handling.

How could AI buyers affect retailers and marketers?
Retailers and marketers may need to optimize for both humans and machines, with more focus on structured data, transparent pricing, and machine-readable product and policy information.

Conclusion

Visa’s research confirms what many in the industry have already sensed: AI-driven commerce is shifting from experimentation to infrastructure. The idea of AI as a customer may still sound futuristic, but the systems behind it are already being built, tested, and deployed. For brands, marketers, and commerce teams, the opportunity now is to adapt before the shift becomes unavoidable. If you want a smarter way to monitor performance and stay agile as digital commerce evolves, ROAS Suite is a practical solution worth considering.