Meta is testing a shopping research feature inside its Meta AI chatbot interface, allowing users to ask product questions and receive carousels of recommended items complete with brand information, pricing, and website references, according to people familiar with the matter.
The feature, which allows requests for product suggestions, is being rolled out to some US based users of the Meta AI web browser. The chatbot responds with a carousel of product images that include captions with information about the brand, website, and price. It also offers a brief explanation of its recommendations in bullet point form.
As per Bloomberg: “The feature, which allows requests for product suggestions, is being rolled out to some U.S.-based users of the Meta AI web browser. The chatbot responds with a carousel of product images that include captions with information about the brand, website, and price. It also offers a brief explanation of its recommendations in bullet-point form.”
The test is a direct response to a shift in how consumers discover and research products. As AI tools have grown more capable of delivering conversational, personalized answers, a growing share of product discovery behaviour that previously flowed through search engines and social feeds has begun migrating toward chatbot interfaces. Meta is attempting to ensure that when its users make that shift, they do so inside Meta AI rather than on a competitor platform.
OpenAI and Google are testing similar functionality, looking to build upon the increasing reliance on AI tools as product research and discovery systems, while also providing revenue opportunities through sponsored listings and advertisements.
The commercial logic is straightforward. Hundreds of billions, and possibly trillions, of dollars are being invested into AI development, and these providers need to demonstrate how their tools are going to make their businesses money, as opposed to simply commanding attention. For Meta specifically, the pressure is acute. The company has committed to spending $600 billion on US based AI infrastructure over the next three years, a figure that demands a credible path to AI driven revenue at scale.
Shopping carousels represent one of the cleaner routes to that revenue. Rather than inserting single sponsored placements into AI responses, which risk making the entire answer feel commercially compromised, carousel displays offer multiple products simultaneously, giving the appearance of choice while still creating monetisation opportunities through paid positioning within the results.
Ads in chatbots could risk eroding trust in AI answers because the responses could be viewed as being skewed toward paying ad partners. Carousel displays of related content can reduce this risk to some degree by giving users more choice instead of highlighting a single paid placement.
The test is currently limited to US users on the Meta AI web interface, and Meta has not confirmed a timeline for broader rollout or detailed exactly how the commercial arrangements behind the product recommendations will work. Whether the shopping feature will extend to Meta AI within WhatsApp, Instagram, and Facebook, where Meta’s audience dwarfs its standalone AI web traffic, will likely become the more consequential question once the US test concludes.
