Facebook Introduces AI Search Mode, Leveraging Public Posts

Meta has added a new 'AI Mode' to Facebook's search functionality. When users perform a search, the 'AI Mode' appears alongside traditional tabs such as 'People' (user search) and 'Marketplace' (buying and selling). Rather than returning a simple list of links, this mode presents AI-generated answers based on information publicly posted on Meta-owned platforms.

The most distinctive feature is that the information sources for search results comprise public content from Meta's entire platform ecosystem, including users' own public posts. This adopts the same approach as the AI search functionality integrated into the Reddit-style forum app that Meta recently announced. Furthermore, users can pose follow-up questions to the initial search results, enabling an interactive format for exploring information in greater depth.

The 'AI Mode' represents just one of several new AI features announced by Meta on the same day. Features such as the 'Photo Preset' function that synthesizes sports uniforms onto fans' photos and collage template suggestion capabilities are also being rolled out simultaneously. Meta is making its strategic direction clear: leveraging its vast social graph and enormous volume of user-generated content as a foundation for AI applications.

This move is driven by intensifying competition in AI search. While Google incorporates AI Overview into search results and Perplexity AI gains prominence through conversational search, Meta is pursuing differentiation by leveraging the 'real-time nature' and 'community information' of social media platforms including Facebook, Instagram, and Threads. Facebook's public posts are particularly rich with information such as local event details and word-of-mouth recommendations that conventional web search struggles to capture, and allowing AI to leverage this information is expected to deliver a unique search experience.

However, concerns regarding transparency and privacy may arise from users' public posts being utilized for AI answer generation. While publicly shared posts are in scope, addressing cases where information is reconstituted and used in ways the poster did not intend will likely become a key focal point moving forward.