Agentic AI: how North America is redefining the grocery shopping journey

AI is entering a new phase in grocery, and its impact is already visible across the path to purchase. After years of experimentation with generative models and machine learning, the industry is now shifting towards agentic AI (systems that act on behalf of the user). This shift is rewriting how shoppers discover, evaluate and buy products, often in ways that sit outside traditional retail environments. For suppliers and retailers, the change is no longer theoretical, it is reshaping shopper behaviour today and accelerating faster in North America than anywhere else.

A new discovery moment

Conversational commerce is one of the clearest signs of change. Shoppers are beginning their shopping on AI platforms rather than on retailer websites, and retailers are pivoting to meet them there.

Target has integrated its full shopping experience into ChatGPT, allowing shoppers to browse, build baskets and check out without leaving the chat. It is also testing retail media placements inside ChatGPT, placing sponsored prompts directly into product‑related conversations. This reflects a broader shift: shopping is increasingly happening beyond traditional touchpoints as third‑party platforms embed commerce into everyday interactions.

Albertsons is piloting AI‑driven advertising that surfaces personalised product suggestions during shopper queries. These moves show how influence is shifting upstream into conversational environments where agents shape intent long before a shopper reaches a digital shelf. This shift creates a new battleground for visibility, while also raising important questions about how retail media will evolve when discovery happens inside an agent rather than on a website.

Checkout without the shopper

The most significant developments are happening at the point of purchase. Google’s new agentic checkout feature allows shoppers to set a desired price for a product and let Google complete the purchase automatically when conditions are met. It is already rolling this out with merchants like Wayfair, Chewy, Quince and select Shopify sellers.

Instacart is moving in the same direction. Its agents can build baskets, compare products, optimise for dietary needs and eventually automate replenishment. This reflects a broader shift towards systems that can receive instructions, create a plan and use tools to complete tasks without requiring explicit inputs.

These developments move decision‑making away from the shopper and towards the agent. When an AI system is optimising for price, nutrition or convenience, brand equity becomes less influential. Suppliers risk losing visibility unless their products are structured and differentiated in ways that agents can interpret. Retailers become interchangeable if AI agents cannot recognise what makes them different.

Retailers build their own ecosystems

Some retailers are building their own agentic platforms to retain control over data, merchandising and shopper relationships. Walmart is leading the charge. Its AI‑powered tools help shoppers find deals, plan events and navigate stores. Shoppers who use Walmart’s app spend 25 percent more per trip than those who do not, showing the commercial upside of AI‑augmented journeys.

In Canada, Loblaw has launched an advanced AI grocery ecosystem, developed in partnership with OpenAI and Google. Its app integrates personalised meal planning, automated basket building, real‑time product comparison and health‑oriented recommendations.

CVS is expanding into this space with its Health100 platform, which uses AI to simplify healthcare‑related shopping through personalised bundles and affordability‑focused recommendations. These examples show how agentic AI is spreading beyond grocery into adjacent missions that influence food, wellness and household purchasing.

Retailers that build their own agents maintain control of the shopper relationship. Those that do not may find themselves disintermediated by Apple, Google, Amazon or OpenAI, whose agents increasingly sit between shoppers and the retail ecosystem.

Retail media enters a new phase

If agents choose products automatically, traditional paid search and sponsored listings become less relevant. Agentic AI will soon start to reshape retail media.

Amazon is already demonstrating what this future could look like. Its AI assistant, Rufus, integrates sponsored prompts directly into the shopping experience and now serves 250 million active customers.

These shifts mean product visibility will increasingly depend on how well items are structured for AI interpretation, while retail media will need to move beyond keyword bidding and into conversational, agent‑driven formats.

What comes next?

A new model of grocery is taking shape, built around conversational discovery, automated checkout, retailer‑built agents and retail media embedded inside AI platforms. This shift will fundamentally reshape how consumers discover, evaluate and purchase products.

For FMCG suppliers and retailers, the priority now is to prepare for a world where agents, not shoppers, make many of the decisions. That means investing in structured product data, clearer differentiation and stronger digital ecosystems. As agents take on more decision‑making, products will need clearer data, stronger cues and more explicit value signals to remain visible in the basket.

Looking for further inspiration?

IGD subscribers can read our report Agentic AI and the future of shopping, which examines how Agentic AI will impact shoppers' path to purchase and what industry needs to do to prepare.

Also check out our exclusive interview with David McIntosh, Chief Connected Stores Officer at Instacart, discussing how agentic AI is changing the industry. Part one here and part two here.