The new feature, called Your Algorithm, allows people to enter their interests in ordinary language rather than relying solely on their viewing habits, reactions and comments to influence the content appearing on their screens.
Meta Platforms announced the rollout on 13 October, saying its recommendation systems could respond to written preferences in real time, including during the same browsing session. The company has not specified a completion date for the rollout or provided a country-by-country availability timetable.
Users can access Your Algorithm through the three-dot menu attached to Facebook posts. The dedicated control lets them add and manage topics they would like the platform to prioritise when selecting recommended posts and short videos.
Facebook is also expanding the feedback available on individual posts and Reels. After someone chooses Interested or Not interested, the platform may ask follow-up questions to establish which aspects of the material appealed to them or prompted their objection.
Those responses can include free-form descriptions of what the user wants to encounter more frequently or avoid. Meta says the additional information will help its ranking systems make finer distinctions between subjects instead of interpreting every positive or negative reaction as a broad preference.
The changes introduce a more conversational method of influencing recommendations, although they do not replace the automated ranking process. Facebook will continue to consider established behavioural signals alongside the preferences users explicitly provide.
Meta said large language models, the technology underlying many generative AI services, make it possible for its systems to interpret the context and intention of written instructions. The models are being incorporated into the mechanisms that determine which content receives prominence.
Previously, Facebook’s recommendation systems inferred interests from activities such as liking posts, leaving comments and watching videos, together with numerous other signals. The new controls add a direct channel through which people can correct or refine those assumptions.
For example, a person who watches a video about a particular sport might want recommendations about that competition without receiving material covering every sport. Written instructions could help the system distinguish between those preferences, although Meta has not published independent performance results for the feature.
The company says adjustments should become visible without requiring users to leave Facebook and return later. That immediate response is a central element of the announcement, but the degree of change may vary according to the content available and other ranking factors.
The revised three-dot menu is intended to make the controls easier to find while browsing, rather than requiring users to navigate through multiple settings pages. The company has not said whether all users will receive the new options simultaneously.
The announcement does not establish that users can switch off personalised recommendations entirely through Your Algorithm. Nor does it describe a facility for viewing or modifying every signal used by Facebook’s ranking systems.
Meta has also not detailed whether written preferences submitted through the new feature will influence advertising recommendations, or how long individual instructions will remain active. The stated functionality concerns the ranking of content in Feed and Reels.
The distinction matters because greater influence over recommended subjects is not the same as complete control over the order or selection of posts. Facebook still determines what qualifies for recommendation and how competing signals are weighed.
The introduction follows similar changes elsewhere within Meta’s services. Instagram introduced Your Algorithm for Reels in December 2025 and subsequently expanded its personalisation options, while Threads has experimented with conversational requests to influence recommended content.
The Facebook version brings that approach to both the main Feed and its short-video service, two prominent places where users encounter posts from accounts and creators beyond those they follow.
Meta’s October announcement did not provide adoption targets, anticipated changes in engagement or figures showing how accurately the system interprets different instructions. Its description of faster, more relevant recommendations therefore remains a company assessment rather than an independently established outcome.
Existing Interested and Not interested selections remain part of the experience, with the additional questions designed to collect more specific explanations when Facebook presents them. The company has not indicated whether every feedback selection will trigger a question or whether users can request clarification independently.
Follow Arabian Post
Select Arabian Post as your preferred source on Google and MSN News for trusted business news and Arab politics and updates.
