Mastercard and Krungthai Card have completed Thailand's first authenticated agentic payment transaction, powered by Mastercard Agent Pay.
Conducted in a controlled pilot environment, the transaction was initiated by an AI agent using Mastercard Agent Pay, with mobility provider Elife serving as the merchant.
In the inaugural test case, an AI agent booked a ride from Suvarnabhumi airport to Central Chidlom via Elife's ground transportation network. The booking and associated payment were handled autonomously by the agent, with tokenised credentials authenticated through Mastercard Payment Passkeys to support strong customer verification and data protection.
How Mastercard Agent Pay works
Mastercard Agent Pay is designed to provide the security infrastructure required for AI-initiated transactions. Each transaction uses a Mastercard Agentic Token issued uniquely to the agent involved. Consumer consent is explicitly captured prior to the transaction, and purchase confirmation is secured via Mastercard Payment Passkeys, keeping authentication and control with the cardholder throughout.
KTC is the first partner in Thailand to test agentic AI transactions with Mastercard. A company official from KTC noted that the collaboration reflects a strategic commitment to integrating agentic commerce into its broader ecosystem.
Regional rollout and infrastructure
The Thailand pilot follows a series of authenticated agentic transactions that Mastercard completed across Asia Pacific earlier in 2026, including deployments in Australia, New Zealand, Singapore, Malaysia, India, South Korea, Taiwan, and Hong Kong. Together, these milestones reflect a phased regional expansion of agent-initiated payment infrastructure.
The intended expansion covers additional industry segments beyond travel, including entertainment and retail, suggesting the technology is being positioned as a broadly applicable payments layer rather than a sector-specific solution.
For the wider payments ecosystem, the Thailand pilot illustrates how tokenisation, passkey authentication, and AI agent orchestration can be combined within existing card network infrastructure. Moreover, the use of established authentication standards, rather than entirely new frameworks, points to a strategy of extending current security protocols into emerging agentic use cases, which may lower the barrier to adoption for financial institutions already operating within Mastercard's network.
