Mastercard and Danske Bank complete Denmark’s first AI agent payment
Mastercard and Danske Bank have completed Denmark’s first payment made entirely by an AI agent on a consumer’s behalf.
The consumer asked the AI agent to book a coffee tasting experience through Priceless.com, Mastercard’s experience platform, after which the agent carried out both the booking and the payment using a Mastercard-branded card issued by Danske Bank.
How the transaction worked
The payment was processed through Mastercard Agent Pay, a framework designed to allow AI agents to initiate and complete purchases while keeping the consumer’s consent and control central to the process. Under this setup, the consumer’s approval is obtained explicitly before a transaction proceeds, and the purchase is confirmed through Mastercard Payment Passkeys, an authentication method intended to verify that the transaction has been actioned by the account holder. PayOS managed the technical execution of the transaction from initiation to completion.
According to Erik Gutwasser, Division President at Mastercard Northern Europe, the framework is intended to ensure that AI-initiated payments remain transparent and authenticated, while also giving card issuers greater visibility into transactions triggered by AI agents rather than directly by the cardholder. Mark Wraa-Hansen, Head of Personal Banking Denmark at Danske Bank, added that the collaboration reflects the bank’s approach to exploring AI-powered payments with an emphasis on customer value, trust, and control.
Context and industry relevance
The transaction forms part of a broader move by Mastercard to extend agentic commerce capabilities across Europe, building on the introduction of Mastercard Agent Pay as a mechanism for securing AI-agent-initiated purchases. As AI agents increasingly take on tasks such as product discovery, booking, and payment execution on behalf of consumers, payment networks, and issuers are developing frameworks to address the authentication and consent challenges this raises. Unlike a conventional card-not-present transaction initiated directly by a consumer, an AI-agent transaction introduces an intermediary layer of decision-making, which raises questions around verification, liability, and fraud prevention that frameworks such as Agent Pay are designed to address through explicit consent capture and passkey-based confirmation.
Plans for the future
Mastercard has indicated that it will continue collaborating with partners across Europe to expand the use of authenticated AI-agent transactions across industries beyond the initial use case demonstrated with Danske Bank in Denmark. The Denmark transaction serves as an early public demonstration of the Agent Pay framework in a live commercial setting, following its introduction as part of Mastercard’s broader agentic commerce strategy. Earlier in 2026, Mastercard completed agentic transactions in other European countries, such as Ukraine, Moldova, and the Netherlands.
In an exclusive comment for The Paypers, Vadim Drozd, CEO of payment infrastructure company FinteqHub, argues that AI-agent payments will first scale in low-value, routine purchases, while more complex transactions will require much stricter user-defined conditions, permissions, and safeguards.
`With its first transaction initiated by an AI agent, Denmark joins a growing number of countries that have already tested interactions with AI agents in real-world payment scenarios since December 2025. Notably, the payment processed by Danske Bank utilized Mastercard’s existing infrastructure. The company will likely strengthen its market position as more banks join its ecosystem. Meanwhile, banks also benefit from boosted transaction volumes and new customer inflow.
By 2027, such payments could account for 5% of total transaction volume, primarily involving recurring, low-value purchases with straightforward terms. Psychologically, consumers find it easier to entrust a program with reordering groceries or paying for a routine service than with booking an expensive vacation. Consequently, adoption will likely begin with specific, individual tasks where budgets and constraints can be pre-set. For merchants, this offers an opportunity to reduce cart abandonment rates during checkout. However, a rise in automated payments doesn’t necessarily imply a commensurate increase in consumer spending, as these agents will initially handle purchases consumers intended to make anyway.`
