Nearly 80 Percent of Polish Companies Open to AI Agent Purchases, Adyen Survey Finds

More and more Poles are letting artificial intelligence search for and choose products on their behalf, and companies are starting to adapt, albeit cautiously. A new Adyen survey shows that nearly 80 percent of businesses in Poland say they are open to agentic commerce, a retail model in which all or part of the purchasing process is carried out by an autonomous AI agent acting on the customer's behalf.

What Is Agentic Commerce

Agentic commerce is a setup where, instead of a person browsing store websites, an AI agent handles the shopping. The user states a goal in natural language, for example asking to find a specific product at a set price with delivery on a chosen day, and the agent independently compares offers, checks availability, and can complete the transaction within an agreed payment limit.

It's more than an AI-powered search engine. The model assumes online stores will need to handle two parallel types of interaction, one aimed at people and the other at machine systems that read product data rather than photos and page layout.

Watching Rather Than Adopting

The Adyen survey found that despite this declared openness, most Polish companies treat agentic commerce as something to watch rather than an urgent implementation project. Only just over a quarter of the surveyed businesses name it as a strategic priority for the coming year, and a similar share report concrete investment plans.

Adyen notes that many stores still treat artificial intelligence as an add-on to their existing marketing rather than redesigning the entire purchase path for systems that will read and compare offers automatically.

Worries Over Customer Relationships

Companies' biggest worry is weaker direct contact with consumers. According to the survey, 38 percent of businesses fear that an AI agent will become an intermediary that reduces customer loyalty to a given brand, narrowing choice down to the lowest price and best logistics regardless of how a store builds its image.

On the other hand, consumers themselves have doubts about the quality of recommendations. A third of respondents cite trust as a key factor, asking outright whether an AI agent actually recommends the best price and best-quality product, or follows other criteria, such as supplier agreements.

Competitive advantage may increasingly stem not just from marketing alone, but also, for example, from visibility across different language models - Matouš Michněvič, CEE country manager at Adyen

Fraud and New Risks

The growth of AI-driven shopping also brings new risks to transaction security. Nearly a quarter of companies report increasingly sophisticated fraud attempts linked to the use of AI tools, and close to 30 percent have already deployed AI mechanisms to detect and limit such attempts.

This shows the two-sided nature of the shift: the same technologies meant to make shopping easier for customers are also being used in fraud attempts, forcing companies to invest in security alongside their investments in supporting shopping agents.

What It Means for Polish Retailers

For Polish e-commerce, the survey's key takeaway is that product data is starting to matter more than a site's visual layer. Stores with clean, structured product descriptions, variants, prices and availability will be easier for AI agents to read than those that rely on elaborate graphics and ad campaigns.

Adyen recommends that companies unify the shopping experience across channels, keep prices and availability consistent, and adopt flexible payment integrations that can handle transactions initiated by both people and autonomous systems. Fragmented data and systems risk losing the direct customer relationship to AI intermediaries that take over control of what gets bought and where.