Facebook is rolling out a significant update to its Reels platform, aiming to make video discovery more personal and social. Announced Tuesday, the changes introduce a smarter recommendation engine designed to surface fresher content faster, helping users discover videos more aligned with their interests.
New social features are also launching, including “friend bubbles” that highlight Reels liked by friends and allow users to jump directly into private chats. This update, which includes AI-powered search suggestions, represents Meta’s latest push to refine its short-form video strategy and boost engagement.
A Smarter and Fresher Feed
At the heart of the update is a revamped recommendation engine. Facebook says it has invested heavily in creating a smarter system that understands user interests more quickly, whether for short clips or longer creator vlogs. The goal is to deliver a more tailored and timely video experience.
The results already appear promising. According to the company, time spent watching videos on Facebook has surged by over 20% year-over-year in the US. This suggests the algorithmic tweaks are successfully capturing and retaining user attention in a fiercely competitive landscape.
A key change is the prioritization of newness. The algorithm now surfaces 50% more Reels that were published on the same day a user is scrolling, according to the company’s announcement. This directly addresses user feedback requesting more recent and culturally relevant content.
Users also gain more explicit control over what they see. The platform is emphasizing tools like the “Not Interested” button and an improved Save feature. Each interaction helps fine-tune recommendations, creating a powerful feedback loop that improves the user’s feed over time.
Making Reels a Shared Experience
Beyond the algorithm, Meta is pushing to make Reels less of a solitary activity. The introduction of “friend bubbles” is a clear move to inject a social layer directly into the viewing experience. This feature harkens back to Facebook’s original mission of connecting people over shared interests.
These bubbles are small, unobtrusive icons of a friend’s profile picture that appear on Reels they have liked. Tapping a bubble instantly opens a private chat, encouraging users to connect over shared interests discovered through video content. It’s a seamless bridge from public discovery to private conversation.
Discovery is also getting an AI-powered boost. On select Reels, users will now see suggested search terms, similar to features on rival platforms. This allows them to dive deeper into a topic they enjoy without ever leaving the Reels player, streamlining the exploration process.
The Algorithm’s Approach to AI Content
With the rise of generative AI, a key question is how platforms handle machine-made media. In an interview, Facebook’s Vice President of Product, Jagjit Chawla, clarified that the system does not inherently distinguish between AI-generated and human-created content.
Instead, the algorithm’s focus remains squarely on user signals. Chawla explained, “If you, as a user, are interested in a piece of content which happens to be AI-generated, the recommendations algorithm will determine that, over time, you are interested in this topic and content.”
This means that if users engage with AI content, they will be shown more of it. Conversely, if they ignore or dislike it, the algorithm will learn to filter it out.
In fact, negative feedback is treated with particular importance. Chawla noted that when a user provides a negative signal, such as clicking “Not Interested,” “when Facebook does receive a negative signal… Facebook will ‘take that a lot more seriously from an algorithm standpoint.'” This gives users significant power to curate their own feeds and avoid what some might consider ‘AI slop’.
A Unified Strategy in the Video Wars
This latest update is not an isolated event but part of a broader, deliberate strategy by Meta to sharpen its competitive edge. It follows the company’s decisive move in June 2025 to consolidate all video uploads on Facebook into the Reels format, eliminating the old distinction between Feed videos and Reels.
That consolidation simplified a fragmented ecosystem for creators and engineers. By unifying its video backend, Meta laid the groundwork for a more cohesive product, making it easier to roll out universal features like these new discovery tools and a streamlined monetization program.
The increased reliance on AI also aligns with other recent initiatives. In August, Meta launched an AI tool to automatically dub Reels into other languages, complete with lip-syncing, to help creators reach global audiences. This shows a deep commitment to leveraging AI for practical creator tools.
This strategy reflects a hard-learned lesson for Meta, which initially struggled to counter the rise of TikTok. By integrating video more deeply rather than spinning it off, as it once considered, Meta is betting on a unified platform experience to retain users and creators.
Ultimately, Meta’s vision is for technology to handle the heavy lifting of content delivery. Chawla articulated this by telling creators to “do your thing, and the responsibility and the heavy lift is on technology that we will figure out. As AI advances, it’s just making our job easier.” This philosophy underscores Meta’s push to automate and optimize the platform, allowing human creativity to remain the focus.


