Agentic commerce remains closer to an early-stage infrastructure buildout than a technology to transform everyday life—or the broader market. Even optimistic forecasts for 2030 suggest it will account for only a small share of total transaction value, making scale, interoperability, and practical integration more important than headline-grabbing announcements.
That gives established payments players, rather than AI companies, an advantage, according to Christopher Miller, Lead Analyst of Emerging Payments at Javelin Strategy & Research. Miller’s new report, Payments Partners With AI: Incumbents Have the Advantage, looks at why payments incumbents are best positioned to bring agentic AI capabilities to financial services.
Vision Surpasses Reality
The maturity of agentic protocols and tools is comparable to that of Gopher, an early-1990s system for accessing files on remote server that preceded the World Wide Web. The comparison underscores how far the technology remains from its broader ambitions.
“It requires steps to get somewhere,” Miller said. “Just because somebody has a new idea and technology enables a vision of that idea, it doesn’t mean that thing is going to fall out of the sky tomorrow. We have to focus on what exists today, what’s being built right now, what’s going to come into existence tomorrow.”
Recent announcements have highlighted efforts to bring agentic capabilities to payment-focused products, often through partnerships between established payments players and AI firms. But these initiatives don’t always amount to the sweeping transformation their publicity suggests.
“Any of the announcements that you see from incumbents have to be read carefully because there’s two sides to them,” Miller said. “Some of these companies are deploying AI capabilities in the products that they sell, while others are using AI products themselves to produce the products that they sell. Those are related, but separate, ultimately. You could clearly use AI to change how you produce things without ever selling something that involved AI in the product itself.”
An Advantage, Not a Guarantee
Some partnerships between incumbents and AI companies are presented as being at the forefront of the AI revolution. In practice, however, they may involve more limited applications, such as allowing Anthropic to modify certain back-office procedures at a financial institution.
Still, established payments players offer the most likely path for deploying agentic protocols at scale. That gives them an advantage, but it doesn’t guarantee that these efforts will succeed.
“Payments or financial services companies are most likely to buy these things from companies they have heard of,” Miller said. “That is the default built-in advantage. But if they end up not working as promised, it’s worth calling out, because we just emerged from an entire generation where innovative new technologies largely came from new companies, not the incumbents.”
Financial institutions should already be assessing the risk profiles of agentic transactions while helping shape the standards that will govern them. Payments firms, card issuers, and other entities that move funds before a transaction is settled will need to determine how to address the risks that arise. Some of that work will require hands-on experimentation; some will depend on research and participation in industry initiatives.
“There are some institutions that simply have no way of being able to do this on their own,” Miller said. “They won’t have access to data on the scale it’s necessary to learn. Others might. Most institutions will use some sort of third-party risk evaluation or validation service, whatever the case might be. For most institutions, it will be a third-party that has the scale and the data that is most accurate and most relevant.”
The Path Ahead for Merchants
For merchants, the immediate implications are different. Agentic shopping is currently more about product discovery than financial transactions. Merchants should focus on understanding where their customers come from and assessing how agentic commerce could affect their existing customer-acquisition channels.
The impact will vary by merchant. Some may see little change in the near term, while those more dependent on channels vulnerable to disruption may need to act sooner. These businesses should pay closer attention to how consumers discover products through emerging AI services and consider how to build brand awareness within them.
As with search engine optimization, the right approach will vary by merchant type, size, and strategy. There will be no single timeline or playbook for adapting to agentic commerce.
“For merchants, at this point, agentic commerce is not even really about payment,” Miller said. “It is about how you will acquire and interact with customers. If that changes substantially, that has really big impact on business model. That is a front of house question. How you manage payments and evaluate risk and whatever, that’s a back of house question. It’s important, but it’s just operational challenges. But how do you get visible with customers if they start looking in different places? That’s existential.”
Take a Deep Breath
Fortunately, merchants still have time before they need to make major strategic decisions about AI. Despite the hype and pressure from AI companies, agentic commerce remains early in its maturity curve, leaving room to assess how the technology develops before committing to sweeping changes.
“There’s a tendency to focus on the unprecedented nature of agentic AI, but while the technical capabilities might be unprecedented, the frameworks for thinking about it are not,” Miller said. “How do you decide which way to go? We have plenty of time to think about it, and anybody who tells you that we don’t is trying to sell you something. Probably their agentic commerce tool.”
