Mastercard & Santander Trial AI 'Agentic Payments'
Mastercard and Santander have completed Europe’s first live, end-to-end payment executed by an AI agent. The milestone in “agentic payments” involves AI systems initiating and completing transactions on behalf of customers, a move aimed at reducing friction in experiences like hands-free shopping. As AI agents act for users, robust digital identity and real-time risk controls become essential.
This European pilot follows the 2025 launch of Mastercard Agent Pay for Citi and US Bank cardholders, with PayPal also integrating the solution to allow AI agents to complete transactions on behalf of its users. The system works by integrating AI agents as governed participants in the payment flow, allowing them to interact securely with issuers, acquirers, and merchants through existing payment networks. The push for agentic payments is happening alongside a major evolution in real-time payment infrastructure. In the U.S., The Clearing House's RTP network, launched in 2017, now reaches institutions holding about 90% of demand deposit accounts and has a transaction cap of $1 million. The Federal Reserve's FedNow service, which went live in July 2023, has seen rapid onboarding of financial institutions, though the RTP network still has a greater reach in terms of accounts covered. As autonomous transactions grow, digital identity becomes a critical foundation for security and compliance with KYC (Know Your Customer) and AML (Anti-Money Laundering) regulations. Innovations in this space include multi-layered fraud prevention that uses biometric liveness detection, deep document forensics, and real-time behavioral risk analysis. Regulatory frameworks like the EU's eIDAS 2.0 are driving the adoption of digital identity wallets, enabling secure cross-border sharing of personal and financial information. AI's role in payments extends beyond agents to significantly enhance fraud detection. Machine learning models can analyze vast datasets to spot anomalies, identify synthetic identity clusters, and even use behavioral biometrics like typing speed to detect bots. Mastercard reports that its AI-powered solutions have boosted fraud detection rates by an average of 20%, with some instances seeing improvements as high as 300%. For product leaders navigating this landscape, influencing without authority is a key skill. This involves earning trust and commitment through clear communication, credibility, and consistently aligning cross-functional teams behind a shared product vision. It requires understanding the motivations and constraints of different stakeholders, from engineering's focus on system stability to sales' targets, to effectively build consensus. The rise of embedded finance, where financial services are integrated into non-financial platforms, is reshaping partnerships between banks and fintechs. This trend allows banks to access new customer bases and create new revenue streams with lower customer acquisition costs. For fintechs, Banking-as-a-Service (BaaS) partnerships provide the necessary infrastructure to offer regulated financial products without needing a full banking license. In the background, institutional adoption of blockchain and stablecoins is creating new real-time settlement rails. Stablecoins like USDC are being used for 24/7 cross-border transfers, bypassing traditional banking hours and delays. While institutional interest in holding crypto assets directly has been somewhat flat, the use of stablecoins for payments and treasury operations is growing, signaling a deeper integration of blockchain technology into the financial system.