Artificial intelligence (AI) is transforming the shopping experience. Customers can now use AI tools such as ChatGPT, Gemini, and Copilot to research options, gather information, and make purchasing decisions. As AI agents become increasingly capable and adoption grows, many businesses will need to rethink how customers interact with their products and services.

It’s important for organizations to understand how AI is reshaping digital commerce and changing customer expectations. In this article, we’ll explore what agentic commerce entails, why it matters for businesses in Germany, and the opportunities and challenges it is likely to present.

Key takeaways

• Agentic commerce is changing how people discover, compare, and buy products and services.
  
• Customers can use AI agents to complete approved tasks on their behalf, while remaining in control of purchasing decisions.
  
• Businesses will want commerce infrastructure that supports secure payments, authorization, and fraud prevention.
  
• Shared Payment Tokens (SPTs) and merchant controls help enable secure agent-assisted transactions.
  
• Agentic commerce adoption is growing in Germany, but long-term success will depend on customer trust, privacy, and secure infrastructure.
  

What is agentic commerce?

Agentic commerce is a model of commerce in which individuals or businesses use AI agents to discover, compare, and assist with purchasing products and services. Rather than simply answering questions or making recommendations, the AI agent completes user-approved tasks across the buying journey, from researching options to initiating checkout.

Unlike traditional ecommerce experiences, agentic commerce is not tied to a single website or application (app). Users might interact with an AI agent through a chat assistant like ChatGPT, a software app like Microsoft Copilot, or an embedded shopping experience. The user might have the agent help plan a budget-friendly trip, compare software solutions, find a product with positive customer reviews, or assist with other purchasing tasks.

Today’s agentic commerce experiences are delegated rather than fully autonomous. People remain in control of purchasing decisions, while AI agents assist with research, comparison, and other approved actions. Depending on the experience, customers might complete checkout directly within an AI agent or continue the purchase on a business’s website.

As agentic commerce becomes a more common way for people to shop, companies will want infrastructure that enables secure interactions between the three parties involved: the business, the customer, and the agent. This includes ensuring payment credentials are shared securely, protecting against fraud, and preserving the relationship with the customer throughout the agent-assisted transaction.

How is agentic commerce changing buying behavior?

AI is changing how people discover, evaluate, and buy products and services. Instead of starting every purchase on a search engine or business website, many customers now use AI tools to research options and compare alternatives, narrowing their choices before visiting a business directly.

This shift is already visible in Germany. Stripe research shows that 37% of German customers already use AI tools for recommendations and product comparisons, while 21% use them to plan and book travel. Further, assuming certain tools were made available in Germany (e.g., Instant Checkout with ChatGPT), 42% of customers said they would consider completing purchases directly with AI.

From search to agent-assisted buying

Traditional ecommerce requires customers to search across multiple websites, compare products, and manually complete each step of the buying journey. With agentic commerce, users can have AI agents complete many of these tasks on their behalf. Rather than searching individual sites, customers can describe what they are looking for, while AI agents identify suitable options, compare offers, and present the most relevant choices.

This shift is not limited to retail. Customers can use AI agents to research travel options, compare software, identify business services, or find the best offer across a wide range of products and services.

What this change means for businesses

As customers shift toward agentic commerce, businesses will want to ensure that AI agents can easily discover and evaluate their products and services, and assist in the purchasing process. Success is likely to depend not only on traditional ecommerce capabilities, but also on commerce infrastructure that enables trusted interactions between the customer, AI agent, and business.

Key challenges of agentic commerce: From trust to security

Trust is going to be extremely important, as AI agents start to become a regular part of the buying journey. Customers need confidence that agents will only perform actions that they, the customer, have authorized. Businesses need to ensure transactions remain secure and compliant. Agentic commerce might change how purchases are initiated, but it does not change the fundamental responsibilities of the businesses involved.

Being the responsible party in an agent-assisted transaction

Even when an AI agent initiates a purchase on behalf of a customer, the business remains the merchant of record (MoR) and is responsible for the transaction. This includes handling refunds, chargebacks, customer support, and other postpurchase processes. Rather than shifting responsibility to AI agents, agentic commerce introduces a new interface through which customers interact with businesses.

Ensuring payments are secure with Shared Payment Tokens (SPTs)

Supporting agent-assisted commerce requires payment credentials that AI agents can use without exposing sensitive payment information. When a customer gives their permission to make a payment using their preferred payment method, the AI agent uses a Shared Payment Token (SPT) to make that payment; the SPT keeps the underlying payment credentials protected. Because payment credentials remain secure, businesses can support agent-assisted transactions without compromising customer trust.

Implementing authorization and fraud prevention mechanisms

AI agents must only be able to perform actions within permissions defined by the customer. Businesses therefore need mechanisms such as approval requirements, spending limits, merchant controls, and time-bound authorizations that clearly define what an agent is allowed to do.

Alongside these controls, fraud prevention remains key. AI agents introduce new interaction patterns, but they do not remove the need to detect suspicious activity or prevent unauthorized transactions. Solutions such as Stripe Radar help businesses identify and block fraudulent activity while allowing legitimate agent-assisted payments to proceed securely.

Building and maintaining user trust

The long-term success of agentic commerce depends on creating trustworthy experiences for both customers and businesses. This requires secure payment credentials, transparent authorization models, effective fraud protection, and clearly defined responsibilities across the entire purchasing journey. Together, these capabilities enable businesses to adopt agentic commerce while maintaining the security and trust customers expect.

Agentic commerce adoption in Germany

Agentic commerce is gaining momentum in Germany. As businesses introduce new AI-powered experiences, adoption is expected to grow steadily. At the same time, businesses must balance innovation with Germany’s strong expectations around privacy, security, and customer control—factors that are likely to shape adoption in the years ahead.

Data protection and customer trust

Trust remains one of the most important factors for agentic commerce adoption. Users expect transparency about how AI agents use their data, what actions they are authorized to perform, and how purchases are approved. As a result, businesses need solutions that combine strong privacy protections with clear authorization and fraud prevention.

Modernizing existing commerce infrastructure

Many German businesses already operate mature ecommerce platforms, payment systems, and customer management tools. Rather than replacing these systems, agentic commerce is likely to build on existing infrastructure by introducing new capabilities that allow AI agents to discover products, interact securely with businesses, and support customer-approved transactions.

Gradual adoption

Agentic commerce is unlikely to replace traditional ecommerce overnight. Instead, businesses are expected to introduce AI-assisted experiences incrementally. Generally, this means starting with product discovery, comparison, and customer support before expanding into more advanced purchasing experiences, as customer trust and supporting infrastructure mature.

FAQs about agentic commerce