Facebook users will be able to shape more of what appears in their feeds by telling the platform which topics they want to see. Meta is bringing its “Your Algorithm” feature to Facebook, where people can add or manage interests and respond to individual posts with feedback.
The change gives users a more direct way to influence recommendations. Meta says updates can take effect right away, while the company is also using large language models more extensively in the systems that rank content.
Topic controls put recommendations in users’ hands
With “Your Algorithm,” Facebook users can manage the topics they want to see in their feed and Facebook Reels. The feature is designed to let people signal both positive and negative preferences, helping guide which subjects appear in recommendations.
Users can also provide feedback from the three dots menu on an individual post. Options to mark something “interested” or “not interested” may appear beneath posts as well, giving users another way to communicate what they want more or less of.
These controls make recommendations less of a one-way process. Instead of only responding to what they watch or engage with, users can state a preference directly and use it to steer what Facebook surfaces.
Feedback can include a written explanation
Meta says Facebook may ask follow-up questions to better understand why a user liked or disliked a topic. Users can also describe in their own words what they would like to see more or less of, and the company says recommendations will update in real time.
That language-based feedback is central to Meta’s description of its AI approach. The company says large language models can account for the interests, context, and intent behind natural language, so people can tell the recommendation system what they want instead of relying only on preset choices.
Meta says recommendations can adapt within the same session. In practice, the company describes a system that can respond to a request about what matters to a user at that moment, then adjust the content it recommends accordingly.
A broader set of Meta recommendation tools
Facebook’s “Your Algorithm” feature follows a version launched on Instagram earlier this year. Meta has also introduced a similar “Dear Algo” feature on Threads, extending the idea of user-directed recommendations across its social platforms.
On Facebook, the feature is part of what Meta calls a set of AI-powered tools for customizing what people see. The company says it is also making greater use of large language models within its ranking systems, connecting conversational feedback with the way content is selected and ordered.
Together, topic controls and written requests offer two ways to communicate preferences: users can select whether a topic interests them, or explain in ordinary language what they would like to change. Meta presents both as ways for recommendations to respond more quickly to a user’s direction.
What the update changes for users
The announced controls give people more explicit influence over recommendations in Facebook feeds and Reels. They can add and manage topics, react to specific posts, and potentially answer follow-up questions or provide written context about their preferences.
Meta says those signals will shape recommendations in real time. The update also reflects a broader shift in how the company describes its ranking systems: users can communicate preferences in natural language, while AI tools interpret that input and adjust what the platform recommends.
The source describes the controls and Meta’s intended behavior, but does not specify when the Facebook feature will be available to all users. Its central promise is a more direct conversation between users and the system deciding what content appears in their recommendations.