Meta launches Facebook feature letting users customise AI recommendations

MUMBAI: Tired of your feed reading you wrong? Meta is giving Facebook users a more direct say in what they see, introducing an AI-powered feature called ‘Your algorithm’ that lets them fine-tune recommendations across Feed and Reels by spelling out their interests in their own words.

The feature creates a dedicated space where users can add and manage preferences, while offering more detailed feedback on individual posts. Accessible through the three-dot menu, it allows users to indicate whether they want to see more of a particular type of content or less of it.

Users can also answer follow-up questions to clarify their preferences, giving Facebook a more specific picture of what interests them. Meta says recommendations can adapt in real time, even within the same session.

Facebook has traditionally relied on signals such as likes, comments and viewing behaviour to predict what users want to see. Its recommendation systems analyse thousands of signals to rank content and personalise individual feeds.

With ‘Your algorithm’, Meta is adding another layer to that process: natural-language input. Large language models (LLMs) will help interpret interests and intent expressed by users, allowing them to communicate preferences rather than relying solely on their online activity to shape recommendations.

The system will combine these explicit instructions with existing behavioural signals. For instance, users can indicate that they want more content around a particular interest or would rather see less of a certain category. Feedback through options such as ‘Interested’ and ‘Not interested’ can prompt further questions to capture more specific preferences.

The change gives users a more direct way to influence their feeds, although Meta has not disclosed how much weight these new inputs will carry compared with existing ranking signals.

For marketers, creators and publishers, the development matters because recommendation systems influence which organic posts and videos reach audiences on Facebook. Changes to personalisation could affect how content is discovered, even as the immediate focus remains on users shaping their own feeds.

If people begin stating their interests more explicitly, Facebook will have another source of information to draw on alongside observed engagement. However, Meta has not quantified the feature’s potential impact on organic reach, engagement or content visibility.

The distinction between content recommendations and advertising is important. The announcement concerns personalisation across Feed and Reels; it does not establish changes to advertising-targeting rules, ad pricing or the way advertisers buy media on Facebook.

Nor has Meta announced a new advertising product or commercial offering linked to ‘Your algorithm’.

The feature marks a shift towards letting users tell Facebook what they want to see, rather than leaving the platform to infer their preferences entirely from what they watch, like or share.

Yet the algorithm remains in the driving seat when it comes to ranking content. Explicit feedback gives users another way to guide recommendations, but the eventual effect on what appears in their feeds will depend on how Meta incorporates those preferences into its systems.

For the advertising and marketing industry, the key question is whether more direct user input will materially change content discovery and audience engagement. For now, Meta’s update offers users a new way to steer their feeds, while leaving its wider commercial implications to be seen.