Meta announced Tuesday local time that it is rolling out a new AI-powered feature for Facebook called “Your Algorithm.” Users can enter their preferences in everyday language, and those preferences will be reflected in recommendations in Feed and Reels. The feature first launched on Instagram and is now expanding to Facebook.
An era of telling recommendations what you want—in plain language
When you open Facebook, the posts you want to see aren’t always what appears. Many people have likely experienced scrolling past something while thinking, “I’ve had enough of this topic.” Until now, recommendations have largely relied on systems inferring users’ preferences from their behavior, such as likes and watch time.
This new “Your Algorithm” feature changes that relationship a little. Users can enter, in natural language, topics they want to see or ones they want to see less of. For example, “I want to see more cooking videos” or “Show me less gossip.” AI reads those instructions and applies them to the posts shown in Feed and Reels.
Meta says changes to settings are reflected in recommendations in real time. The core of the feature is that users can directly state what they want instead of leaving everything to automated recommendations.
Express your preferences right from the three-dot menu
Another notable change is an update to the three-dot menu shown on individual posts. Opening the menu lets users add the post to “Your Algorithm” or choose “Interested” or “Not interested.” It’s designed to let people indicate their preferences right when they come across a post that catches their attention.
Not having to hunt through settings is a major difference for users. Social media settings are often buried deep, and many people keep using the service without knowing they exist. If a post right in front of them becomes the entry point, they’ll naturally have more opportunities to use the feature.
Follow-up questions probe why users like or dislike something
Meta says it may show follow-up questions to better understand what users liked or disliked. Simply tapping “Not interested” doesn’t tell the system whether the content itself was a poor fit or whether the user didn’t like the creator. Getting users to explain why can provide clues for improving the accuracy of adjustments.
From the user’s perspective, this adds a little extra effort. But if a few seconds spent answering can make the feed more comfortable to browse afterward, it may be worth trying.
Social media recommendations are moving from passive to interactive
This move builds on Meta’s efforts to let users customize recommendations. On Instagram, it launched a feature with the same name for Reels in December 2025, and expanded it to Feed and the Explore tab in June 2026. Threads also has a related feature called “Dear Algo,” which lets users influence the recommendations they see. With the rollout to Facebook, the same approach is now taking shape across Meta’s major services.
The backdrop is the growing role AI plays in content discovery. Traditional recommendations have mostly inferred preferences behind the scenes from behavioral data. This feature adds room for users to express their wishes in words. In terms of being able to adjust what you like yourself, it’s a move toward an experience more like that of a video streaming service.
What it means to be able to express your preferences in words
That said, this doesn’t mean the inner workings of the algorithm are becoming visible. The platform still controls the framework for how posts are arranged. What this feature expands is the way users can say what they want to see.
For people accustomed to chatbots, expressing preferences as if talking to AI may feel natural. On the other hand, how much your daily feed changes to match your tastes is something you’ll only know by trying it. After using it for a few days, see whether the feel of your feed has changed.
Comments