Agentic AI for shopping journeys: a market snapshot of emerging innovations

There are countless reports and industry examples of how companies and retailers are using AI to automate their processes. Now the next revolution is on the horizon: not only retailers will be able to use AI in this way, but consumers too.

A new generation of AI might be stepping directly into the role of the shopper. This is the promise of agentic AI: autonomous digital agents that can not only search, inform, compare, decide, and plan, but also act in the real world! In a few years, agentic AI systems in retail could be executing purchases, making payments, booking services, and managing transactions on behalf of the user. This could automate the entire purchasing journey. Agentic AI can initiate an order, fill in personal data, complete checkouts, and even handle post-purchase tasks like tracking or reordering.

The race to build, integrate, or power this new layer of commerce is heating up. In just the past few months, from April to June 2025, several major players like Walmart, Amazon, Visa, Mastercard, PayPal and Perplexity have announced their moves in the realm of agentic commerce.

Google is pushing the transition to agent-based commerce together with leading commerce platforms, retailers, and payment providers, as the company announced in January 2026. With the Universal Commerce Protocol (UCP), Google introduced an open standard that connects AI agents, merchants, platforms, and payment services across the entire customer journey – from product discovery and checkout to after-sales. Development partners include Shopify, Etsy, Wayfair, Target, and Walmart, supported by players such as Adyen, Mastercard, Visa, Stripe, American Express, and Zalando.

In practical terms, UCP will enable direct checkouts on Google’s AI interfaces, such as AI Mode in Search and the Gemini app, including Google Pay and, in the future, PayPal – while merchants retain full control as seller of record. In addition, Google is introducing the Business Agent, a brand-specific AI sales assistant within Search that provides advice, presents offers, and, over time, will also enable agent-driven direct purchases. New Merchant Center data attributes and the Direct Offers ad format are designed to help merchants remain visible in conversational commerce and activate customers who are ready to buy.

Microsoft is expanding its retail portfolio with agentic AI solutions such as Copilot Checkout, which enables purchases directly from AI-powered conversations – without redirecting users to external shops. Merchants remain merchant of record. The AI checkout is available in the US and is supported by partners including Shopify, PayPal, and Stripe. Early adopters include Urban Outfitters, Anthropologie, Ashley Furniture, and Etsy sellers.

In addition, Microsoft is introducing agentic shopping solutions for personalized customer experiences: Brand Agents that bring brand voice and product expertise directly into digital touchpoints, as well as a customizable shopping agent template in Copilot Studio. This enables conversational product recommendations, contextual advice, and advanced features such as outfit creation across web, mobile, and in-store channels.

Amazon’s “Buy For Me” feature introduces agentic AI introduces agentic AI into its shopping experience by enabling customers to purchase items from third-party brand stores directly through the Amazon app if those products aren’t currently available on Amazon. The agentic system automates the checkout process on external sites by securely transmitting encrypted customer data and their payment information to complete the purchase. The selection and confirmation of the purchase is still carried out by the consumers themselves, while the actual purchase process on the third-party website is fully automated by AI.

One of Amazon's goals here, in addition to an even more convenient customer journey: Elevating the marketplace from being merely a product discovery platform to a centralized agentic purchasing hub, ensuring users never need to leave its ecosystem while still accessing a broader range of products.

Perplexity, a conversational search engine, has partnered with PayPal to embed agentic commerce into its user experience. Starting in the U.S., users of the Perplexity Pro platform will be able to not only research and compare products in natural language but also make direct purchases like buying products or tickets or booking travel. Consumers will be able to purchase products, buy tickets or book travel within the chat flow using payment service providers PayPal or Venmo. The entire transaction process – including checkout, shipping, tracking, and invoicing – will happen using tokenized wallets and passkeys and the user’s PayPal account. This transforms Perplexity from a smart Q&A assistant into a one-stop conversational commerce interface.

“Sparky” is a generative AI-powered assistant integrated into the Walmart app, helping customers with tasks like product search, summarizing reviews, and making occasion-based recommendations. Launched with a focus on enhancing the decision-making phase of the shopper journey, Sparky provides real-time answers and supports informed purchasing.

In June, Walmart announced, it is actively developing Sparky into a fully agentic AI system. Planned capabilities include multimodal interaction (text, image, audio, video), autonomous reordering, service bookings, and end-to-end task handling. With agentic AI, a simple question like “What’s for dinner?” will trigger the agentic AI to generate a week of family-approved meal plans, automatically add all the necessary ingredients to your cart and purchasing them.

Visa’s “Intelligent Commerce” initiative marks its entry into agentic retail by offering the backend tools needed to empower AI agents to handle the entire customer shopping journey – including payment. Rather than creating a direct-to-consumer solution, Visa provides APIs and SDKs for third parties to integrate secure payment features like tokenization, authentication, and passkey checkout. Visa’s plan is to position itself as foundational infrastructure for the emerging AI-powered shopping world, where the agent – not the human – is the shopper. Their partnerships with major players like OpenAI, Anthropic, IBM, Microsoft, Stripe, and others reinforce Visa’s commitment to facilitating secure and scalable agentic commerce.

Accenture is investing in retail tech company Profitmind via Accenture Ventures while also entering into a strategic partnership to further embed agentic AI in retail. The goal is to help retailers automate operational decision-making in areas such as pricing, inventory management, promotions, and planning, and to move more quickly from insight to execution.

Profitmind relies on a network of specialized AI agents that consolidate diverse data sources and derive prioritized, actionable recommendations. The platform combines machine learning, generative AI, and agentic AI to accelerate decision-making – while maintaining transparency around the logic and expected impact of AI-driven recommendations.

SAP announced in January 2026 that it will support retailers in agentic commerce with concrete, ready-to-deploy solutions. The Storefront Model Context Protocol (MCP) Server for SAP Commerce Cloud is expected to be available from Q2. It is designed to enable “channel-less commerce” by allowing retailers to interact securely with a wide range of AI agents – both within their own touchpoints and via external assistants such as ChatGPT or Perplexity. The MCP server supports product presentation and purchasing across channels for both humans and machines and is part of SAP’s broader strategy to support agentic standards such as MCP, ACP, and UCP.

In an agent-driven commerce environment, product data must be understandable, structured, and trustworthy for machines. With the Catalog Optimization Agent in SAP Commerce Cloud, SAP automates the maintenance and enrichment of product catalogs at scale. This is intended to help retailers better position their content and products for AI-based purchasing decisions.

Agentic AI marks a turning point in the evolution of digital commerce, promising consumers faster, more intuitive, and more hands-off shopping experiences. By offloading tasks like searching, comparing, and even checking out, it offers unprecedented convenience. But it also raises questions around trust, transparency, data privacy, and how much control users are willing to delegate.

For retailers, the shift to agent-driven shopping opens new opportunities to reach customers in more seamless, embedded ways – yet it also challenges existing interfaces, brand visibility, and loyalty strategies. As agentic commerce grows, retailers need to rethink how they can compete for ever cheaper deals and faster deliveries, which USPs they need to develop, and how they can maintain relationships with their customers.