Polish mobile payment network BLIK has completed its first pilot agentic payment transaction, allowing an AI agent to monitor a product and complete the purchase automatically once it becomes available.

The pilot was carried out with Polish self-care brand Your KAYA, technology startup Juo, and payments provider PayU. An AI assistant on the Your KAYA online store was instructed to monitor an out-of-stock product: a light hand cream with coconut and almond fragrance. When the selected variant returned, the AI agent added it to the basket and completed the purchase using BLIK.

The transaction was worth PLN 19.99. Most importantly, the user did not need to intervene when the purchase was finally made. The payment happened under conditions the customer had approved earlier through their banking app.

BLIK Brings Agentic Commerce Into Payments

The pilot offers a practical example of how agentic commerce could change online shopping.

Until now, AI shopping assistants have largely focused on helping consumers find products, compare options, and make recommendations. Payment has often remained the final step requiring direct customer involvement.

BLIK’s pilot connects those two parts.

The customer first tells the AI agent what to look for and gives it permission to act within defined conditions. The agent then monitors the merchant’s inventory. Once the requested product becomes available, it can complete the purchase without asking the customer to return to the checkout screen.

That creates a different shopping model.

Instead of the consumer searching, deciding, and paying at the same moment, an AI agent can potentially manage the entire process based on instructions provided earlier.

For merchants, this could open new possibilities for product alerts, subscriptions, replenishment, and other purchases where timing matters.

Payment Consent Remains at the Center

The most important part of the BLIK pilot may be the way consumer control is built into the process.

BLIK says the agent could not simply decide to spend money on its own. The user first had to provide an instruction and approve payment through their banking application.

The customer therefore defined the boundaries within which the agent could operate.

That distinction is critical as AI agents become more capable.

Giving an agent access to payment infrastructure creates a new risk question: how much authority should the agent have?

BLIK’s model points toward a system in which the consumer gives advance consent rather than approving every individual transaction.

The agent can then execute a purchase only within those previously established conditions.

For financial institutions and payment companies, this type of permission architecture could become increasingly important as consumers begin delegating more shopping tasks to AI.

How the BLIK Transaction Worked

The pilot involved several layers of technology.

The customer started a conversation with an AI assistant inside the Your KAYA online store. The desired product was unavailable at the time, so the customer instructed the agent to continue monitoring the store.

The agent repeatedly checked availability.

When the selected product came back into stock, it detected the change, added the product to the shopping basket, and completed the purchase.

The payment was processed through PayU, while Juo provided the technology enabling the AI-agent conversation and agentic commerce functionality. BLIK supplied the payment mechanism.

The process demonstrates how agentic payments are likely to work in practice.

No single company has to build the entire system. The AI layer, merchant experience, and payment infrastructure can operate together.

That could make agentic commerce easier to deploy across existing payment ecosystems.

Why This Matters for Payments

Payments have traditionally been built around direct customer actions.

A consumer chooses a product, enters payment information, or authorizes a payment, and the transaction is completed.

Agentic commerce changes the role of the customer.

The consumer may give instructions once and allow software to perform subsequent actions on their behalf.

That could make payments more convenient, particularly for routine or time-sensitive purchases.

Consider products that frequently sell out, recurring household items, or services where the best price changes quickly. An AI agent could monitor these conditions and act when the customer’s requirements are met.

The payment system then needs to recognize not just the transaction itself but also the authority under which the agent is acting.

That is where BLIK’s pilot becomes relevant beyond one purchase.

BLIK Tests a New Type of Customer Experience

For Your KAYA, the pilot provides a glimpse into how AI could become part of e-commerce rather than simply an additional customer service channel.

Customers could eventually delegate a wider range of tasks, from finding products to managing replenishment or recurring purchases.

Juo’s involvement is also significant because the Polish startup is focused on subscription infrastructure and agentic commerce.

The company says its technology is designed to support flexible subscription models, recurring transaction logic, and agent-ready e-commerce experiences.

That makes the pilot more than a one off experiment.

It combines several trends that are developing at the same time: AI assistants, automated purchasing, recurring payments, and digital wallets.

As those technologies converge, the payment itself becomes almost invisible to the customer.

The instruction becomes the primary interaction.

PayU Adds the Payment Layer

PayU’s role demonstrates another important part of the emerging ecosystem.

Payment providers will need to support transactions initiated by software while still maintaining security, authentication, and consumer protection.

In a conventional checkout flow, the customer is visibly present.

With agentic commerce, the customer may not be present when the actual payment is triggered.

That means the payment infrastructure needs to understand whether the transaction is authorized, whether it fits the customer’s predefined conditions, and whether the agent is acting within its permitted scope.

PayU’s involvement in the BLIK pilot shows how payment processors could become an important bridge between AI agents and existing merchant infrastructure.

BLIK’s Broader Digital Payment Position

The timing is notable for BLIK itself.

The payment network reported 1.6 billion transactions in the first half of 2026, with 21.6 million active users at the end of June.

The system was also classified by Poland’s National Bank of Poland (NBP) as a systemically important retail payment system on September 9, reflecting its growing role in the country’s payments market.

The agentic-commerce pilot therefore comes as BLIK is expanding both its scale and the range of use cases it supports.

For BLIK, allowing an AI agent to initiate a purchase is a natural extension of that development.

The bigger ambition is to make the payment method useful in new digital experiences without compromising the security and trust that underpin the system.

The Key Question: How Much Should an AI Be Allowed to Spend?

The BLIK transaction also raises an issue that the payments industry will need to solve as agentic commerce expands.

Consumers may be comfortable allowing an agent to buy a low-cost product such as a hand cream.

The situation becomes more complicated when the purchase is expensive.

Would customers allow an agent to purchase a flight, book a hotel, reorder business supplies, or make a financial transaction without immediate approval?

The answer will depend heavily on controls.

Spending limits, merchant restrictions, product requirements, time windows, and transaction notifications could all become important parts of agentic payment systems.

BLIK’s pilot points toward one possible model: the consumer remains in control by defining the rules in advance, while the agent handles execution.

That could provide a balance between convenience and oversight.

Agentic Payments Move From Concept to Real Transactions

Much of the discussion around AI agents in payments has so far focused on future possibilities.

BLIK’s pilot is important because it demonstrates an actual transaction.

The agent did not merely recommend a product or prepare a basket. It monitored availability, detected the requested item, purchased it, and completed the payment.

The user was not required to intervene at the moment of checkout because the authorization had already been established.

That makes the transaction a useful early example of what agentic payments could look like in real commerce.

The technology is still at an early stage, and a single pilot does not prove that consumers are ready to hand broad purchasing authority to AI.

However, it does demonstrate that the basic components can work together.

TheFinRate Take

BLIK’s first agentic payment transaction is significant because it moves AI-powered shopping from product discovery into actual payment execution.

The PLN 19.99 purchase may be small, but the model behind it is much bigger.

A customer gave an AI agent a task, established payment consent through their bank, and allowed the agent to act when predefined conditions were met. Juo provided the agentic layer, PayU processed the payment, and BLIK supplied the payment method.

That is the architecture the wider payments industry will be watching.

The future of agentic commerce will not depend only on whether AI can find the right product. It will depend on whether consumers trust agents to spend money on their behalf—and whether payment networks can provide the controls to make that delegation safe.

BLIK’s pilot suggests that the next step in digital payments may not be faster checkout. It may be checkout that happens without the customer being there at all.

First pilot BLIK transaction completed independently by an AI agent