Santander and Mastercard say they have completed Europe’s first live end-to-end AI agent payment.
The payment was executed by an AI agent, with the two banks describing it as the first agentic payment carried out within a regulated banking framework.
An AI agent payment is designed to remove a moment of effort, and that can change the customer journey far beyond checkout.
In discussing the news, Matías Sánchez, Global Head of Cards and Digital Solutions at Santander said:
“At Santander, we see AI as a transformative force in the evolution of payments. Our role is not only to adopt innovation, but to shape it responsibly, embedding security, governance and customer protection by design.”
“As AI agents become part of everyday commerce, building trusted, scalable frameworks will be essential to unlocking their full potential.”
What Santander And Mastercard Actually Tested
Santander carried out the transaction in a controlled environment using Mastercard Agent Pay.
The companies say Santander processed the transaction through its live payments infrastructure to validate the end-to-end operational and control framework under real conditions.
The solution enables AI agents to initiate and execute payments on behalf of customers within predefined limits and permissions. The approach also maintains strict standards of security, privacy, and consumer protection while using existing payment networks.
Mastercard says Mastercard Agent Pay integrates AI agents into the payment flow as visible, governed participants. Adding that PayOS supported the end-to-end orchestration of the transaction.
Why This Strengthens The Case That Finance Leads
In a recent CX Today article, we cited Lloyds Banking Group as an enterprise blueprint for agentic AI adoption, highlighting that generative AI delivered around £50 million of value for Lloyds in 2025, and over 50 AI use cases went live that year.
Also Lloyds expects £100 million in incremental value in 2026 alone from its next wave of agentic AI, with income verification for mortgages shifting from days to seconds, while fraud and dispute workflows that once stretched across weeks are now resolved in days.
The earlier article also pointed to Mastercard’s broader bet on agentic AI. Mastercard believes a third of enterprise software applications could incorporate agentic AI by 2028, and it expects a significant percentage of customer interactions and operational tasks to be supported by AI agents by 2030.
This new Santander and Mastercard pilot extends the storyline into payments. It suggests agentic AI is not only taking on service and operational tasks, it is also starting to execute transactions.
And innovation is unlikely to stop here, Kelly Devine, President, Europe at Mastercard said:
“Agentic payments represent a profound shift in how commerce is initiated and executed. This milestone with Banco Santander demonstrates that innovation and trust can advance together.”
Where Friction Could Move Next In The Customer Journey
Checkout has been a long-running battleground for digital teams. Faster flows helped, and wallets helped too.
Agentic payments aim to remove the moment of customer effort altogether. If customers can delegate the purchase, friction reduction becomes less about screen design and more about delegation design.
Discovery May Become More Automated
An agent can compare options quickly, and it can do it consistently. Brands may need to compete on what the agent can verify, including availability, policies, and clean product data.
Consent Becomes The New Front Door
The new friction point is permissioning. Customers will care about what an agent can buy, when it can buy, and what it must ask before spending.
They will also care about how easily they can review, reverse, or dispute an outcome. A customer might let an agent reorder household basics within a monthly limit, while a procurement team might let an agent buy approved items from approved merchants.
Support And Disputes Start To Look Different
As agents take on simpler steps, exceptions become the main event. Returns, refunds, and disputes could become agent-led, and that could reduce customer effort.
But it could also expose policy gaps faster, and it could raise new questions about intent. Service teams may need new playbooks, and they will be helping a person who owns the outcome and an agent that executed the action.
What CX Leaders Should Do Next
This is still early, but it is a useful planning signal. Start by identifying which journey steps customers want to delegate, and which steps still require direct confirmation.
Then involve the risk owners early. CX, payments, fraud, compliance, and legal teams will shape what permissions mean, and what proof is required when something goes wrong.
This does not constitute a commercial rollout at this stage. Santander says it will move into extended testing and scaling, exploring additional use cases and partnerships while maintaining strong controls, resilience, and regulatory alignment.
If AI agents start buying on our behalf, the journey becomes quieter on the surface. But expectations get louder, customers will demand speed and proof that the system stayed within their intent.
The next advantage may be simple: when a customer sends an agent to buy, does your business make trust easy?
Join the conversation: Join our LinkedIn community (40,000+ members): https://www.linkedin.com/groups/1951190/
Get the weekly rundown: Subscribe to our newsletter: http://cxtoday.com/sign-up
