Meta launched AI search mode on Facebook: answers from posts, groups, and Reels

Facebook is turning years of user posts into a live search knowledge base. Meta has launched AI Mode — a new search feature powered by Meta AI that aggregates answers directly from public content across the platform.

What is AI Mode

AI Mode is a fundamentally new way to search for information on Facebook. Instead of a classic list of links and posts based on keywords, users receive a summary answer compiled by Meta AI from public content: group discussions, Reels, Marketplace listings, and public posts from pages. Similar to how Google works with AI Overviews in Search, Facebook synthesizes answers based on multiple sources within the social network. The key difference is that the sources here are not indexed web pages, but real conversations and experiences from actual people.

The feature has begun rolling out to users in the US in stages. The new function allows, for example, asking "where are the best pizzerias in Brooklyn" and getting a summary from hundreds of discussions in local groups — rather than just a list of business pages. This is exactly what users have long wanted from a social network that, in terms of user knowledge volume, has long surpassed most specialized forums and review sites.

What is included in the search

AI Mode covers several types of public content:

• Facebook Groups — discussions, recommendations, and reviews from topical communities
• Reels — video content, which is now indexed as a source of answers to queries
• Marketplace — listings for selling goods, rentals, and services
• Public posts from users and business pages
• Comments in open groups and discussions

This means that if someone wrote a detailed review in the "Best Restaurants in Barcelona" group or posted a Reel with an honest review of a new smartphone — this data can be included in the answer to someone's search query. Facebook is turning accumulated user UGC into a structured knowledge base accessible through natural language.

Why Meta needs this

Facebook has accumulated nearly twenty years of user content: billions of reviews, recommendations, discussions, and video reviews. All of this previously remained difficult to access — finding what you needed was only possible through standard search, which often delivered mediocre results. AI Mode changes the logic. The platform analyzes the meaning of the query and synthesizes an answer from multiple public sources.

This is Meta's direct response to several competitive threats: TikTok with video recommendations for younger audiences, Google with AI Overviews, ChatGPT and Perplexity, which are siphoning information traffic away from traditional search engines. Notably, Generation Z is increasingly using social networks as search engines — and Meta clearly wants to reclaim that traffic.

Meta is turning the social network into a search engine built on trust

in the opinions of real people — not just indexed web pages.

For Meta, this is also a monetization opportunity: search traffic within the platform can be monetized through ads inserted into AI answers — the same way Google has long earned from Search Ads.

What this means

AI Mode is a signal that traditional search within social networks is becoming a thing of the past. Facebook is transforming from a news feed and messenger into a space where AI processes years of real conversations and delivers structured answers. For businesses and content creators, the rules are changing: public posts in groups and Reels now affect not only reach in the feed but also what users discover through search. Public presence in communities becomes a new form of SEO.

There's a separate question about privacy. For now, AI Mode only works with public content, but users should keep in mind: what they wrote in open groups is now actively used in the platform's search results.

*Meta is recognized as an extremist organization and is banned in Russia.