Visa and Mastercard Launch Agentic AI Tools to Accelerate Payment Automation

Visa is expanding beyond its role as a credit and debit card transaction network into value-added merchant services, focusing on agentic AI, stablecoins and digital assets. The shift could determine whether Visa can reduce its reliance on card-processing fees as payment methods rapidly evolve, while generating revenue from fraud prevention, data analytics and new settlement tools.

Visa executives recently said the company is advancing agentic commerce, in which AI agents can independently search, shop and make payments. It is also using AI to detect and block payment fraud such as romance scams while developing stablecoin applications. Reports did not provide investment figures, a formal launch date or projected new revenue, and Visa has yet to disclose a specific timetable.

Visa is a major global payments network, while OpenAI develops generative AI and AI agent technology. Their strategic alliance aims to enable AI agents not only to process information but also to autonomously complete payments and financial transactions under preset rules, creating trusted payment infrastructure for automated commerce.

Visa has unveiled Programmable Money Rails and is integrating OpenAI technology to advance the AI agent economy. The framework is expected to allow agents to initiate and execute payments autonomously. Reports have not disclosed a formal launch date, the value of the partnership, transaction limits or the first institutions set to adopt the system.

AI agents are evolving from providing recommendations to searching, making decisions and paying on consumers’ behalf. The shift requires the payments industry to address authorization, identity verification, risk controls and consumer control. Building on Visa Intelligent Commerce, Visa Agentic Ready allows issuers to test agentic transactions over existing card networks, moving the concept into controlled, real-world payment flows.

Visa launched the program in Europe on March 17, 2026, with 21 issuers participating, including Barclays, HSBC UK and Revolut. It expanded to Latin America and the Caribbean on April 29, followed by 10 Asia-Pacific markets with more than 50 partners on April 30. Taiwan’s initial group includes nine issuers, among them Cathay United Bank, CTBC Bank and E.SUN Bank. Visa did not disclose an investment amount.

Agentic commerce allows software to search for, select and purchase products with a user’s authorization, making identity verification, payment authorization and risk controls critical. Visa launched Visa Intelligent Commerce on April 30, 2025, to provide payment APIs to AI platforms. Connect further integrates the connection process, lowering adoption barriers for merchants, agent developers and payment service providers.

Visa unveiled Intelligent Commerce Connect on April 8, 2026. Through a single integration with the Visa Acceptance Platform, businesses can initiate payments, tokenize credentials, set spending controls and perform authentication, with support for APIs from Visa and other card networks. Seven companies, including AWS, Mesh and Payabli, are piloting the service, with a broader rollout planned during 2026. Visa did not disclose pricing or an investment amount.

AI agents are moving beyond recommendations to make purchases, call APIs and buy computing resources, but existing checkout systems are still designed for humans. Stripe and payments blockchain Tempo therefore jointly developed the open Machine Payments Protocol, or MPP, allowing agents to receive payment requirements, authorize payments and access services in a single request. Visa’s integration with existing card networks is significant because it brings machine micropayments into mainstream payment and risk-control systems.

Stripe and Tempo released the MPP specification on March 18, 2026, alongside the launch of Tempo’s mainnet. Visa released a card-based specification and SDK the same day and enabled them on the Visa Acceptance Platform, with support for tokenization, authentication and agent identity verification. Visa said on July 14 that MPP had settled about $25,000 across roughly 115,000 transactions in the weeks since its mid-March launch, averaging less than 1 cent per transaction.

Agentic commerce allows AI agents, with consumers’ authorization, to search for products, place orders and make payments. Visa and Banco Santander integrated Visa Intelligent Commerce (VIC) with existing payment infrastructure to test consent capture, data security and interoperability across merchants and payment networks. A Visa survey found that more than 70% of Latin American consumers already use AI during the shopping process, making transaction controls and consumer protection critical to scaling the technology.

Visa and Banco Santander announced on March 12, 2026, that they had completed Latin America’s first end-to-end AI agent payment pilots across five markets: Argentina, Brazil, Chile, Mexico and Uruguay. An AI agent bought chocolate in Brazil and books in the other four countries, completing the payments within existing regulatory and authorization parameters. The companies did not disclose transaction values or the number of participants.

AI agent payments allow systems not only to recommend products but also to complete transactions autonomously within spending limits and authorizations set by users. The significance of the collaboration between Banco Santander and Mastercard lies in bringing this form of agentic commerce into a regulated banking framework and testing whether existing payment networks can maintain security, privacy and consumer protection.

On March 2, 2026, Banco Santander used Mastercard Agent Pay in a controlled environment, with PayOS providing end-to-end orchestration, to complete Europe’s first live transaction initiated and executed by an AI agent through Santander’s existing payment infrastructure. The companies did not disclose the transaction value and stressed that the project remains a pilot and has not been commercially launched. They plan to expand testing and explore additional use cases.