Meta tests an AI shopping assistant, challenging ChatGPT and Gemini

The AI assistant wars have entered territory where serious money circulates — e-commerce. Meta Platforms is testing a product research feature in its Meta AI chatbot, aiming to provide direct competition to similar shopping tools already offered by OpenAI's ChatGPT and Google's Gemini. Bloomberg agency reported this information, citing its own sources.

At first glance, the news may seem routine: another company adding another feature to another chatbot. But looking deeper, it becomes clear why shopping functionality specifically has become a battlefield for the largest tech corporations. The global e-commerce market is valued in the trillions of dollars, and whoever manages to capture the moment of purchase decision — when a user chooses between three headphone models or compares smartphone specifications — gains control over a colossal advertising and partnership stream.

For Meta, this move looks particularly organic. The company already owns powerful commercial infrastructure: Facebook Marketplace is used by hundreds of millions monthly, Instagram Shopping has turned the social network into a storefront for brands, and WhatsApp in some countries has become a full-fledged sales channel. Until now, however, these tools existed separately from the company's AI assistant. Integrating the shopping feature into Meta AI could become the connecting link that closes the cycle: a user sees a product in their Instagram feed, asks the chatbot about its specifications and alternatives, receives a personalized recommendation, and makes a purchase — without leaving Meta's ecosystem.

Competitors are not standing still. In early 2026, OpenAI is actively developing ChatGPT's commercial capabilities, adding product comparison features and integration with major retailers. Google has gone even further, combining Gemini's capabilities with Google Shopping and search advertising, essentially creating an AI layer on top of its traditional advertising machine. Each player has their own trump cards: OpenAI bets on dialogue quality and depth of analysis, Google — on the scale of search data and advertising infrastructure. Meta's trump card is its social graph and preference data from nearly four billion users of its platforms.

Technically, the task of creating a quality shopping assistant is significantly more complex than it appears. A chatbot must not simply find products — it must understand the context of the query, account for the user's budget, compare specifications from different sources, distinguish advertising descriptions from real reviews, and meanwhile remain neutral, not turning into an advertising tool for specific brands. The last point is particularly important and especially problematic for Meta, whose business model has historically been built on advertising. How to reconcile objective AI assistant recommendations with the interests of advertisers who pay for promoting their products? This question still has no elegant answer from any of the companies.

There is also another important aspect — user trust. Research shows that people still approach AI recommendations with a certain skepticism, especially when it comes to spending. Companies will have to convince their audience that their AI assistant really is looking for the best option for the buyer, not for the advertiser. In conditions where Meta has repeatedly found itself at the center of scandals related to user data processing, this task becomes even more difficult.

The consequences for the market could be enormous. If the largest AI platforms successfully implement shopping features, this will hit traditional product aggregators, review sites, and even search engines in their classical form. Why open ten tabs and read reviews if you can ask one question to an AI assistant and get a structured answer? Amazon, Alibaba, and other marketplaces are also developing their own AI assistants, but they lack the social context that Meta possesses.

The race for the user's wallet through the AI interface is only beginning. Meta enters it late, but with a serious set of advantages. The question is whether the company will be able to convert its social data and commercial infrastructure into a truly useful tool — or whether the shopping feature will become just another channel for showing ads in new packaging. The answer to this question will determine not only the fate of Meta AI, but also how billions of people will make purchases in the coming years.