Meta starts testing its AI shopping assistant

Engadget

by Mariella Moon

March 3, 2026

AI-Generated Deep Dive Summary

Meta has quietly begun testing an AI-driven shopping assistant for select users in the U.S., initially accessible only on desktop browsers via Meta AI’s website. The tool appears when users see the “Shopping research” button within the query box. Upon asking for product recommendations, the chatbot displays a carousel of images, prices, links to e-commerce sites, brand info, and brief explanations for its suggestions. If Meta AI can access user data like gender or location, it tailors responses—for example, suggesting women’s puffer jackets available in New York based on the tester’s profile. While users can’t checkout directly from the interface, they can click through to shop online. This marks Meta’s latest move into agentic shopping tools, which Zuckerberg previewed earlier this year during an earnings call. While the company is entering a crowded space, with competitors like OpenAI (via ChatGPT), Google (with Gemini), and Perplexity already offering similar AI-powered shopping assistants, Meta’s tool aims to differentiate itself through integration with its ecosystem and tailored recommendations based on user data. The development aligns with broader trends in AI-driven retail solutions, designed to enhance user experience by streamlining product discovery. While the current iteration is limited in scope, it signals Meta’s ambition to leverage its AI capabilities to compete in the e-commerce space. As more companies invest in AI shopping tools, consumers can expect increasingly personalized and efficient shopping experiences.

Verticals

techconsumer-tech

Originally published on Engadget on 3/3/2026