Meta Turns Facebook’s Public Posts Into AI Search Answers
Facebook users can now ask its artificial intelligence questions and get answers drawn from the public posts of other users, part of a new AI Mode that Meta introduced this week. The feature lets the social network’s assistant search across public content, summarize it and answer questions, while the company also rolled out AI-powered photo editing and content creation tools. Analysts estimate the search product could generate as much as $10 billion in annual revenue, a figure that reflects how much value Meta believes it can extract from the vast archive of public posts its users have created over two decades.
The launch turns Facebook’s social graph into a data moat. Meta has spent years building one of the largest collections of user-generated content in existence, and AI Mode puts that collection to work in a way the company has not attempted before. Instead of sending users to a search engine or an answer service, Meta’s assistant answers directly from what people have already posted publicly, stitching together opinions, experiences and information into responses. For Meta, the appeal is obvious: no other company has access to the same reservoir of real-world, person-generated content.
The privacy questions are equally obvious. The posts being used as answers are public, but many users posted them without imagining they would become raw material for an AI assistant, or searchable in a new way. The feature has reopened the debate that has shadowed every AI company: what consent means when people’s words become training and retrieval material. Meta said the feature only draws on posts that are already public and that it respects the account settings users have chosen, but critics said the framing misses the point, since most users do not understand what public posts are now feeding.
There is also the matter of what the answers are worth. The $10 billion revenue estimate comes from analysts who believe AI search can monetize through advertising, subscriptions and commerce, the same mechanisms that already fund Facebook’s core business. If users come to the platform to get answers, the theory goes, they will see ads and services around those answers, and Meta can charge a premium for the traffic. The estimate is a projection, not a forecast, and skeptics note that monetizing AI-generated answers has proven difficult for every company that has tried, since users who get an answer may not linger long enough to see an ad.
The launch comes on a day that also brought a notable departure. Emily Dalton Smith, the executive who led Meta’s “AI for Work” initiative, has left the company, according to people familiar with the matter, ending a tenure focused on bringing AI tools to the enterprise. Meta did not comment on the departure, which follows a pattern of leadership changes as the company reorganizes around its AI ambitions. The timing is awkward, coming on the same day the company announced a consumer product meant to define its AI direction, and it has fed speculation about internal disagreements over strategy.
Meta’s broader AI push has been a story of heavy spending and uneven results. The company has invested tens of billions in computing infrastructure, built its own large language models, and integrated AI into its apps and advertising systems. Some of those efforts have produced results, particularly in ad targeting and content recommendation, where AI has directly improved revenue. Others, including Meta’s AI assistant products, have yet to demonstrate the usage and monetization that would justify the investment. AI Mode is an attempt to connect the two tracks: use the assistant to drive engagement, and use engagement to drive the advertising revenue that is the company’s core business.
The competitive context sharpens the stakes. Rivals have shown that social content can power successful AI products, and the major platforms are racing to integrate assistants into their services. Meta’s advantage is the scale and specificity of its content, which skews toward the personal, the local and the lived experience in ways that interest advertisers. The company’s challenge is to present that advantage without triggering the backlash that has followed every expansion of how user data is used. Its history, including the scandals that reshaped its approach to privacy, makes that a delicate balance.
For users, the practical effect of AI Mode is that their public posts have acquired a new function they did not choose. What was written to share with friends is now retrievable by strangers through an assistant, in forms the original author never saw. Meta has offered no way to opt individual posts out of the system beyond making them private, and the company argues the feature improves the value of its platform. Whether users agree will show up in engagement data and in the regulatory attention the feature draws. Meta has bet that the answers are worth more than the questions. This article was prepared by Rhino Finance’s editorial team based on public reporting.
