Banco Santander and Mastercard have announced the completion of one of Europe's first live end-to-end payments executed by an artificial intelligence (AI) agent, conducted within a regulated banking framework.
The transaction was carried out using Mastercard's Agent Pay solution and processed through Santander's live payments infrastructure under controlled conditions.
The test marks what both companies describe as a milestone in the deployment of agentic payment systems, AI-driven models in which software agents initiate and complete transactions on behalf of customers, operating within predefined limits and permissions. The pilot was conducted within Santander's regulated payment framework and does not constitute a commercial rollout at this stage.
How the transaction worked
Mastercard Agent Pay integrates AI agents into the payment flow as visible, governed participants, enabling interaction between issuers, acquirers, and merchants. PayOS supported the end-to-end orchestration of the transaction. The solution is designed to allow an AI system to complete a purchase using existing payment networks while maintaining controls around security, privacy, and consumer protection.
The transaction was processed through Santander's live infrastructure specifically to validate the operational and control framework under real conditions, rather than in a simulated environment.
Matías Sánchez, Global Head of Cards and Digital Solutions at Santander, noted that the bank views AI as a significant force in the evolution of payments, and that its approach centres on embedding security, governance, and customer protection into the design of any such framework. Adding to this, Kelly Devine, President, Europe at Mastercard, stated that agentic payments represent a shift in how commerce is initiated and executed, and that the principles underpinning its network, including security, trust, interoperability, and scale, are being applied to AI-enabled commerce.
Next steps and regulatory context
Following the pilot, Santander will move into extended testing and scaling, exploring additional use cases and partnerships. The bank has indicated it will maintain regulatory alignment and resilience controls throughout this process.
The development sits within a broader industry conversation around the governance of AI agents operating in financial services. As regulators across Europe continue to assess frameworks for AI deployment in banking, the Santander–Mastercard pilot represents one of the first publicly confirmed instances of a regulated institution processing a live agentic payment through operational infrastructure.
Both companies have indicated their intent to continue developing the framework, with the longer-term aim of supporting AI-driven transaction models at scale across existing payment networks.
