Instagram Tests New Feature That Lets Users Customise Their Reels Algorithm

Instagram has begun testing a new feature that allows users to better personalise their algorithm-based recommendations by selecting or removing topics of interest, beginning with Reels and later extending to the Explore page. The update, announced by Instagram chief Adam Mosseri on Threads, is designed to offer users more influence over their feed and browsing experience. The move aligns with Meta's wider push to boost customisation and transparency across its apps, following recent updates that allow users to restrict sensitive content and refine suggested posts.

Instagram Lets a Few Users 'Tune' Their Algorithm

Instagram CEO Adam Mosseri has revealed that the platform is testing a new feature allowing users to “tune their algorithm.” The update is said to be designed to give people greater control over the content that appears in their feeds, helping create a more personalised and engaging browsing experience.

The feature, found under a new Your Algorithm section in Settings, allows users to view and manage the topics influencing their recommendations. It begins by summarising their current engagement patterns, which, in Mosseri's case, include Luxury Watches, Fashion Week, Bad Bunny, Stand-up Comedy, and Concerts.

Users can tap the Add+ button to include more topics they enjoy, or select specific areas they'd like to see less of. Needless to say, this option will only be available once access to the feature is rolled out to a user's account.

Initially, the algorithm customisation option will be available for Reels, enabling participants in the test to fine-tune the videos shown in their feeds. The update builds on Instagram's broader effort to increase user control, following earlier introductions of parental tools, sensitive content limits, and improved content filtering.

Mosseri also confirmed that the feature will later extend to the Explore tab, with long-term plans to introduce similar functionality on Threads. However, he did not specify the number of users involved in the test or provide a timeline for a full rollout, instead encouraging participants to share feedback that could guide future refinements.