Meta Platforms Inc. is trialling a new shopping research feature in its AI chatbot, aiming to rival similar tools from OpenAI’s ChatGPT and Google’s Gemini.
The feature, currently available to select US users via the Meta AI web browser, allows users to request product recommendations. The chatbot responds with a carousel of product images, accompanied by captions detailing brand, website, and price, along with brief bullet-point explanations for each suggestion. A Meta spokesperson confirmed the test but provided no further details.
CEO Mark Zuckerberg has outlined plans to build “personal superintelligence,” positioning Meta’s chatbot to compete with other AI platforms that have begun integrating e-commerce features. In a January earnings call, he said new products will launch soon, offering a “uniquely personal experience” based on users’ history, interests, content, and social connections.
Bloomberg News testing found that the chatbot tailors suggestions according to the user’s location and inferred gender. For instance, searches for puffer jackets prompted recommendations suited for women in New York. While the tool does not include a checkout option, users can follow merchant links to purchase items.
Meta has not clarified whether it receives referral commissions or favours brands advertising on its platforms. However, Zuckerberg previously indicated that these AI shopping tools aim to help users discover specific products from businesses in Meta’s catalog, complementing its targeted advertising ecosystem.
