Walmart and Google push AI agents into online shopping

Walmart and Google are accelerating efforts to embed artificial intelligence agents into online shopping, signalling a shift that could alter how consumers search, compare and buy products on the internet. The two companies say the technology is designed to move beyond traditional search results and recommendation engines, allowing software systems to act directly on a shopper’s behalf.

The retail group has confirmed that it is integrating Google’s Gemini models into parts of its digital commerce operations, including tools that help customers find products, compare options and manage repeat purchases. Executives describe the move as part of a wider industry transition from keyword-based searches to conversational and task-oriented AI systems that can understand intent and complete actions end to end.

Walmart’s leadership has framed the change as a response to evolving consumer behaviour. Shoppers increasingly expect faster, more personalised experiences across apps and websites, particularly on mobile devices. AI agents, in theory, can handle complex requests such as “restock my weekly groceries within this budget” or “find a laptop with these specifications and deliver it by Friday”, without requiring users to click through dozens of listings.

Google, for its part, is positioning Gemini as a foundation for what it calls agentic commerce. Rather than simply surfacing links or ads, Gemini-powered agents are being developed to plan tasks, compare prices across merchants, apply loyalty benefits and complete checkouts when authorised. For Google, whose core business has long depended on search advertising, the pivot reflects an attempt to stay central to online commerce as generative AI changes how people interact with the web.

Industry analysts say the partnership underscores a broader competitive dynamic. Large retailers and technology platforms are racing to define standards for AI-driven shopping before smaller rivals or new entrants can capture consumer trust. Amazon has been developing its own AI shopping assistants, while startups are experimenting with independent agents that can roam multiple marketplaces to find the best deals for users.

Walmart’s scale gives it an advantage in testing such systems. With millions of daily customers and a vast catalogue spanning groceries, electronics and household goods, the company can train and refine AI models on diverse purchasing patterns. It has already used machine learning extensively for inventory management, demand forecasting and pricing, and executives argue that customer-facing agents are a logical next step.

The integration with Gemini is also notable because it suggests a closer alignment between a major retailer and a technology provider that does not operate its own large-scale consumer marketplace. Unlike Amazon, Google does not sell products directly, which reduces conflicts of interest and may make retailers more comfortable relying on its AI infrastructure. At the same time, Google gains access to real-world commerce data that can help improve its models.

There are, however, unresolved questions about control, transparency and consumer protection. If an AI agent is authorised to make purchases autonomously, regulators and consumer advocates are likely to scrutinise how decisions are made, which products are prioritised and whether commercial incentives influence outcomes. Clear disclosures about pricing, substitutions and data usage will be essential to maintain trust.

Privacy is another concern. AI shopping agents require detailed knowledge of user preferences, spending habits and sometimes location data to function effectively. Both companies have said safeguards are being built to ensure users remain in control, with explicit permissions required before any transaction is completed. Still, the accumulation of sensitive data raises the stakes for cybersecurity and governance.

For advertisers and brands, the rise of agent-driven shopping could disrupt established marketing strategies. If consumers rely on AI agents to select products based on constraints such as price, sustainability or delivery speed, traditional sponsored listings may carry less weight. Brands may need to optimise for machine decision-making, ensuring accurate product data and competitive pricing rather than eye-catching ads alone.