Rise of Agentic Commerce: Why Payments Now Demand Layered and AI-Driven Security

We are entering the era of the "Invisible Checkout." For decades, the act of payment has been a conscious human friction, the physical exchange of cash, the swipe of a card, the tap of a phone, or the click of a 'Buy Now' button․ But as Artificial Intelligence evolves from generative chatbots into autonomous agents, the human role is shifting from manual execution toward strategic delegation. Today, AI agents are beginning to negotiate prices, manage subscriptions, and optimize supply chains with minimal human intervention. This shift poses a fundamental question for the global financial architecture: When a machine pays a machine, how do we guarantee human intent?

Oleksandr Yablunivskyy, Vice President and Head of Products and Solutions for Visa CISSEE, says that Visa is moving beyond simple transaction processing to build a "Trust Layer" for the autonomous economy, leveraging tokenization and the new Trusted Agent Protocol to ensure that while payments may become invisible, they never become unauthorized.

-Mr. Yablunivskyy, as AI agents move from providing information to executing tasks, the payment step often becomes invisible to the human user. How is Visa evolving its security and “Proof of Intent” protocols to ensure that when an AI agent pays for something, it is truly what the human consumer authorized?

At Visa, we have spent decades building the infrastructure that underpins trust in commerce, connecting over 14,500 financial institutions and 175 million merchants, and processing hundreds of billions of transactions each year. Agentic commerce is a natural extension of that mission.

In traditional commerce the flow is straightforward: user, merchant, network, issuer. One interaction, one decision. In agentic commerce you now have multiple handshakes, multiple identities, multiple layers of trust, and that changes everything. Checkout is no longer the control point.

As payments become more autonomous, ensuring “proof of intent” becomes critical. Our approach is to embed that intent directly into the transaction flow. With Visa Intelligent Commerce and the Trusted Agent Protocol, consumers define the rules upfront: spending limits, merchant categories, and permitted use cases, and any action taken must operate strictly within those parameters. The Trusted Agent Protocol enables approved AI agents to securely share verified information with merchants, ensuring that each transaction is recognized as authorized and aligned with the consumer’s intent, while also distinguishing trusted actors from malicious automation.

The technical foundation behind all of this is tokenization. Instead of exposing raw card credentials, transactions are enabled through secure, authenticated tokens. With more than 16 billion tokens issued globally by Visa, this infrastructure ensures that even if credentials are compromised, they are effectively unusable, providing a critical layer of protection in an increasingly automated ecosystem.

-In an ecosystem where Agentic AI can negotiate prices or switch payment methods in real-time to optimize for rewards or low fees, how do you see the traditional relationship between banks and customers changing?

-As agentic AI becomes more embedded in commerce, the bank–customer relationship is shifting from transactional to increasingly intent-driven. Customers are no longer just initiating payments; they are delegating decisions to intelligent agents operating within parameters they define.

In this environment, banks’ role evolves toward enabling and safeguarding that delegation. This means providing the infrastructure for secure authorization, clear controls, and trusted identity frameworks, while supporting real-time, data-driven decision-making behind the scenes. Financial institutions will need to align vision with execution by modernizing their technology stacks and embracing AI-driven capabilities to stay relevant in a more autonomous commerce landscape.

At the same time, relationships become more personalized and continuous. With AI acting on behalf of customers, banks have an opportunity to engage earlier in the journey, helping shape preferences, manage risk dynamically, and deliver value through intelligent insights, adaptive credit, and personalized experiences rather than only at the point of payment.

Ultimately, Agentic AI doesn't sideline banks — it elevates them into orchestrators of intelligent finance, embedding their products directly into AI-driven decision flows where speed, trust, and compliance matter most. Research from Deloitte estimates agentic commerce could unlock up to $17.5 trillion in global economic value by 2030, creating massive new demand for bank-powered rails, liquidity, identity, and risk services that AI agents must rely on. In this future, the winners aren't banks with the loudest brands, but those that become the default financial operating system for AI.

-In light of these shifts, what fundamental role will banks play in the new ecosystem, and what are the primary risks inherent in this process?

-In the agentic era, banks remain foundational, with their role becoming more complex and more critical. Financial institutions will need to modernize their infrastructure, embrace AI-driven capabilities, and ensure seamless, secure interoperability across increasingly digital and agent-driven commerce environments.

As AI begins to initiate and optimize transactions, banks effectively become the control layer, ensuring every action is secure, authorized, and aligned with customer intent. This includes strong authentication, clear spending controls, dynamic risk assessment, and maintaining trust at scale. At the same time, banks have a growing opportunity to deliver more personalized, data-driven financial services throughout the entire customer journey.

The risk landscape, however, is evolving just as quickly. Fraud is shifting from stolen credentials to compromised identity: deepfakes, synthetic profiles, and manipulated authorization signals. This requires a fundamentally different response: layered security combining tokenization, biometrics, and real-time AI-driven monitoring. At Visa, for example, we already deploy over 100 AI models to help detect and prevent fraud.

Ultimately, success will depend on how effectively banks balance innovation with trust, advancing their capabilities while reinforcing the safeguards that underpin the entire payments ecosystem.  

-From your perspective, what is one major global trend in technology or business that will shape the next decade?

-If I had to name one trend that will define the next decade, it's the intersection of generative AI and economic digitalization. 

One insight from our 2026 Global Economic Outlook stood out to me. Small businesses are now outpacing individual consumers in adopting generative AI tools. That's a structural shift. We're moving toward a world where a lean team of five people, properly augmented by AI, can operate at the scale and reach of a company ten times its size, fundamentally reshaping competition, entrepreneurship, and commerce. 

The data reinforces just how far this is already progressing. ECDB's Global E-Commerce Trends report shows AI shopping traffic already converts 54% better than traditional sources, and by 2028, AI-driven interfaces are projected to overtake classic search, with around 60% of all searches already zero-click. We are moving rapidly from “AI as a discovery” to AI as the checkout layer itself.

The broader economic picture supports this momentum. Global GDP is projected to grow 2.7% in 2026, which sounds steady, but beneath that stability, the structure of the global economy is being fundamentally rewired. Businesses are accelerating investments in AI infrastructure, supply chains are being restructured, and cross-border payments are recovering faster than domestic ones. These are connected signals of an economy reorganizing around new technological capabilities. The payments ecosystem must evolve to support this new reality, through greater interoperability, real-time capabilities, stronger trust frameworks, and the integration of AI-driven experiences.

Altogether, this means that organizations that invest today in adaptability, digital infrastructure, and AI readiness will be best positioned to actively shape the future of commerce. That’s why this moment in payments is especially exciting.

-What is your primary Call to Action for the innovators and leaders shaping the future of finance. If they want to build this Agentic AI ecosystem rather than just react to it, what should be their first move tomorrow morning?

-My primary call to action is to prioritize trust alongside capability from the outset.  Tomorrow morning, the first move should be to get concrete about how your organization will enable secure, governed interaction between AI agents and payments, defining clear rules for authorization, embedding strong identity and risk controls, and ensuring your systems can support real-time, data-driven decisions at scale. Those who act early to align technology, governance, and partnerships around this foundation will be able to shape how the ecosystem develops.