Meta challenges Google with new AI shopping search feature

Meta Platforms Inc. is testing a shopping research feature in the Meta AI web browser for some US users. It returns product carousels with brand, website, and price information, Bloomberg reported.

Mark Zuckerberg, chief executive officer, said on an earnings call in January that the company wants to build personal superintelligence and will start shipping new products to offer a uniquely personal experience.

Users can click merchant links for further browsing, and Bloomberg found the tool sometimes tailors recommendations to the user’s location and the gender it infers from their name.

A Meta spokesperson confirmed the test but did not respond to questions about whether Meta receives referral commissions.

🔗 Source: Bloomberg

🧠 Food for thought

Implications, context, and why it matters.

The chatbot is the public face of Meta’s agentic commerce strategy

Meta frames the feature as part of CEO Mark Zuckerberg’s “agentic shopping,” where AI finds specific products for users 1.
The plan also ties into end-to-end ad automation. Brands set a budget plus an objective, then the AI runs the workflow 2.
Meta has mentioned “Manus” in its AI monetization plans. The provided sources do not confirm an acquisition of an AI startup called Manus for about $2 billion 1.
In tests in Mexico and the Philippines, early business AI tools drive over one million weekly conversations 1.

Advertisers may gain sales but lose control to Meta’s AI

Meta AI on the web surfaces product carousels with brand, site, and price. Checkout stays outside the chat, so users follow merchant links 3.
Meta has not said if it earns referral fees from those links. It declined to say whether the chatbot boosts brands that advertise on Facebook or Instagram 3.
Meta links conversion lifts to models such as GEM. After GEM rolled out earlier in 2025, Meta says Q2 results rose up to 5% on Instagram and 3% on Facebook Feed 3.
Brands using AI tools like Advantage+ can get stronger results, yet they get less detail on targeting data 4.
If the assistant becomes a main product search route, product placement may lean further on Meta’s algorithm. Marketing teams may spend more time tuning product data plus creative assets to fit the AI 4.