Our experts and industry insiders blog the latest news, studies and current events from inside the credit card industry. Our articles follow strict editorial guidelines.
Key Takeaways
- Credit card issuers may need to rethink strategies to optimize performance in an environment where cardholders increasingly allow AI agents to shop on their behalf.
- Detection systems that aim to stop credit card fraud from human actors must look for different signs to prevent rogue AI agents from making unauthorized purchases.
Many of the biggest players in the credit card industry have agentic commerce on their radars as the first quarter of 2026 pushes on.
Artificial intelligence agents that identify and purchase products on a consumer’s behalf stand to bring a great deal of convenience to shoppers short on time. But AI agents in the marketplace may also compel credit card issuers to shift strategies to protect revenue streams.
Issuers and cardholders may have had safety concerns in the past about providing AI agents access to card credentials as well as purchasing power. But card networks have developed ways to protect the integrity of the process.
“We believe that we are well positioned to be the infrastructure provider and key enabler in agentic commerce so that every agent interaction is trusted and secure,” Ryan McInerney, CEO of Visa, said in a recent conference call with analysts.
When an emerging technology threatens to disrupt longstanding methods of doing business, some companies may struggle to understand how to bring their operations into alignment with the technology.
We caught up with leaders in the business and payments space to dive deeper into agentic commerce and what it may mean to the credit card ecosystem as it evolves.
Certain fintechs may be able to develop solutions for agentic commerce faster than traditional card issuers.
John McDonnell, Chief Strategy Officer at payments platform Wink, told us that the core innovation with the technology isn’t simply that an AI agent can buy products. Instead, it’s that the ecosystem has the ability in real time to prove that an agent should be making a particular purchase under conditions that buyers and issuers have agreed to.
“The biggest revenue unlock is coupling payments with verified identity signals,” McDonnell said. “When an issuer can bind a transaction to a trusted user and a trusted agent permission, approvals go up, disputes go down, and that creates room for new pricing models tied to risk reduction.”
Of course, traditional card issuers aren’t the only entities seeking to take advantage of how the payments marketplace may operate as AI agents that shop on behalf of consumers become more common.
“Fintechs built natively around AI decisioning will likely innovate faster than legacy institutions,” Tim Tynan, CEO of Chargeback Gurus, told us. “If traditional issuers fail to keep up, fintechs could capture a larger share of consumers by offering new features powered by agentic AI.”
A New Set of Fraud Risks Emerge
Prominent companies in the payments space — including leading networks such as Visa and Mastercard — have developed solutions that bring security and trust to agentic commerce.
After all, neither consumers nor businesses in the credit card industry want to see AI agents make purchases they don’t have the authorization to complete.
Having robust controls in place that prevent fraud in agentic commerce will be a key factor in giving cardholders peace of mind regarding the use of an agent to shop on their behalf. Merchants also need to trust that the agent buying from them is acting in alignment with a cardholder’s wishes.
Traditional fraud systems that look for consumer patterns in shopping behaviors aren’t likely to be nearly as effective with AI agents that function differently from people.
Developing an ideal system for eliminating fraud from agentic commerce may take time as stakeholders interact with AI agents more and seek to understand how bad actors can manipulate agents to make unauthorized purchases.
Leaders in the payments arena such as Mastercard and Visa have worked to bring security and trust to agentic commerce.
Tynan told us that cardholders can dispute any charge they haven’t authorized. And if every transaction an AI agent makes can turn into a chargeback — without recourse for the merchant — then the risk for merchants stands to rise sharply.
“On the other hand, if consumers are liable for any purchase their authorized AI agent makes — even in error — they’ll be hesitant to use the technology at all,” Tynan added. “The situation demands careful balancing of consumer and merchant interests, and that’s something the payments industry is still figuring out.”
Chandler Fang is the Co-Founder of t54, a company aiming to bring trust to the agentic economy. Fang told us that while today’s fraud revolves around impersonating people, it will soon become about impersonating agents, hijacking their mandates, or changing the information agents access to make decisions.
“You could have a rogue agent posing as a legitimate purchasing agent or a compromised agent making transactions that technically fall within its authorization but serve a malicious actor,” Fang said. “That requires continuous behavioral risk assessment — not just point-of-transaction checks.”
The Bottom Line
Exactly what percentage of consumers will soon begin using AI agents to regularly shop for them remains to be seen. One recent study projects that agentic commerce may drive as much as $1 trillion in U.S. business-to-consumer retail revenue by 2030.
But credit card issuers should start preparing now for that change if they haven’t already.
