Facebook Revamps Reels Algorithm with AI to Deliver Fresher, More Relevant Content

Facebook is rolling out a major upgrade to its Reels algorithm, aiming to make user experiences more dynamic, timely, and personalised. The new update, announced by Jagjit Chawla, Facebook’s Vice President of Product, marks another big step in Meta’s push to use artificial intelligence to improve what people see on their feeds.

With video watch time on Facebook rising by 20 percent year over year, Meta is rethinking how it connects users with relevant and fresh content. The updated Reels experience will now prioritise showing users 50% more videos posted on the same day they browse, giving a stronger emphasis to timeliness and trending topics.

According to Chawla, the change is driven by user feedback demanding more recent and meaningful content. “Our internal testing shows that this approach increases engagement,” he explained, adding that the goal is to make Reels feel more immediate and relatable.

Alongside the algorithmic refresh, Facebook is introducing AI-powered search suggestions within Reels—similar to the feature that appears at the bottom of TikTok videos. These suggestions will help users discover new creators and topics based on their previous interactions and viewing habits.

Another new feature, called “friend bubbles,” aims to make the experience more social. This allows users to send quick direct messages to friends who have liked a particular Reel, making it easier to spark conversations and interactions around shared content. Currently, there is no separate feed for friends’ Reels, so this feature bridges that gap by encouraging spontaneous engagement within the app.

However, the growing influence of AI-generated content has raised fresh concerns about authenticity, misinformation, and copyright. With tools like OpenAI’s and Midjourney’s generative models making it easier to create lifelike videos, Meta is working to help users distinguish between human and AI-produced content. The company offers both manual and automatic labeling options for AI-generated posts, though Meta admits that detection systems aren’t foolproof.

Chawla emphasised that Facebook’s recommendation engine treats AI-generated and human-created videos equally. “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,” he said. On the other hand, when users mark something as “Not Interested,” that signal has a powerful influence on what appears in their future feeds.

He also highlighted the evolving nature of content formats and how AI simplifies the process for both creators and users. “Technology will obviate the lines between whether it was long or short [form video] or horizontal or live, or any of that,” Chawla said. “When I talk to both sides, users and creators, I [say], do your thing, and the responsibility and the heavy lift is on technology that we will figure out.”

As Meta deepens its reliance on AI, the company’s focus remains clear—empowering creators to produce authentic, high-quality videos while ensuring that users get the most engaging and relevant content on their feeds. Video, Chawla affirmed, will continue to be at the heart of Facebook’s growth strategy.