Mastercard Completes Korea’s First Live Agentic Transactions, Unlocking Trusted AI-Powered Commerce

WHY THIS MATTERS: The successful execution of an autonomous, AI-driven purchase marks a pivotal moment for the future of digital payments. This development transcends a simple geographic expansion; it establishes a critical trust layer—Mastercard Agent Pay—that is essential for scaling agentic commerce. As Large Language Models (LLMs) evolve from merely providing information to executing complex tasks, the financial industry faces the challenge of ensuring security and authentication in a non-human-initiated transaction environment. This pilot in South Korea, following earlier rollouts across APAC and the US, demonstrates a working solution for secure interoperability between the AI agent, the merchant, and the card network. It assures that as consumers delegate their spending to automated systems, the fundamental pillars of fraud prevention and compliance remain intact. This technical framework is what will unlock the next trillion-dollar market of invisible, background commerce.

Mastercard announced the completion of South Korea’s first live, authenticated agentic transaction, a major step in enabling commerce powered by trusted artificial intelligence (AI). 

A new era of AI-powered payments 

Agentic commerce refers to an ecosystem where AI-powered agents search for, discover and make purchases on behalf of cardholders. In this milestone in Korea, an AI agent completed a live ride-booking purchase through a global mobility provider hoppa, leveraging Mastercard AI payments infrastructure.

The AI agent, facilitated by CardInfoLink and connected to hoppa’s platform, automatically searched for available transportation options, booked a ride, and securely paid for a car service from Incheon International Airport to a hotel in Gwanghwamun, Seoul. 

This end-to-end use case was powered by Mastercard Agent Pay, a framework that enables secure, authenticated payments initiated by AI agents. The framework allows seamless interoperability between issuers, acquirers and merchants. By complying with industry standards, Mastercard ensures that cardholders and businesses benefit from trusted, consistent agent-powered experiences across various partners and platforms.

This latest milestone in South Korea follows successful Mastercard Agent Pay launches in the U.S. in 2025, followed by launches in Australia, New Zealand, Singapore, Malaysia and India in 2026.

Building the future of trusted AI commerce

Following this successful transaction, Mastercard plans to accelerate the expansion of AI-enabled commerce ecosystem in South Korea by collaborating closely with card issuers and digital partners. In addition, Mastercard aims to vitalize commercialization of agent-driven payments across various other sectors in the future in South Korea. 

To reinforce its AI leadership across Asia-Pacific region, Mastercard is also strengthening its strategic initiatives, including:

Establishing a regional AI Centre of Excellence in Singapore to serve as a hub for driving technological innovation and reinforcing governance.  

Deepening partnerships with leading large language model (LLM) providers and AI agent developers across the region.  

Jae-Youl Yang, Vice President of Digital Payments, Mastercard Korea, said, “This successful live transaction represents a significant milestone for the future of AI-powered commerce. It demonstrates how AI agents can safely and reliably conduct payments in a real-world commercial environment. Mastercard will continue working closely with our financial institutions, digital partners and merchants to build a trusted ecosystem for AI-driven transactions in Korea.”

FF NEWS TAKE: This absolutely moves the needle. By providing the underlying secure rail for agentic commerce, Mastercard is proactively solving the trust gap before AI-powered transactions become ubiquitous. The key takeaway is the race for standardization. We should watch closely to see which high-value sectors, beyond travel, become the next proving grounds—specifically, subscription management and complex supply chain procurement. Success will depend less on the AI’s intelligence and more on the payment network’s ability to manage risk and provide guaranteed interoperability across disparate agent platforms.