Instagram is extending its Your Algorithm tool to the main Feed, giving users a way to see and edit some of the topics the app believes they want to see.
The feature had already been available for Reels and Explore. Its expansion to Feed matters because Feed remains one of Instagram’s most central surfaces: it is where posts from followed accounts, recommendations, ads, creator content, and habit-forming scrolling all meet.
According to Instagram head Adam Mosseri, Your Algorithm lets people view the topics Instagram has associated with their interests, add topics they want more of, and remove ones they do not want shaping recommendations. For now, the controls are topic-based. Instagram says it is working on broader controls that could eventually include people, moods or “vibes,” and content types.
What Actually Changes
Until now, most Instagram personalization happened indirectly. If a user watched cooking videos, paused on fitness posts, liked a travel Reel, or repeatedly ignored a certain kind of content, the system used those signals to adjust what appeared next. That feedback loop could be useful, but it was mostly invisible.
Your Algorithm makes part of that loop readable. Instead of only inferring from behavior, Instagram is giving users a menu of preference labels they can change themselves.
That is a meaningful shift, even if it is not full control. Users are not being handed the ranking system, the ad system, or a complete switchboard for every post they see. They are getting a more explicit way to correct Instagram when it has overlearned from a passing interest or misunderstood a pattern.
Why Instagram Can Do This Now
Mosseri connected the feature to changes in AI models. Older recommendation systems could identify patterns at scale, but the internal groupings were not necessarily easy for a person to understand. A system might know that one set of videos performed similarly for a user without having a clean human-readable label for that cluster.
Large language models make it easier to turn clusters of related content into plain-language topics. That is what allows Instagram to show a user something closer to: “home workouts,” “streetwear,” or “travel photography,” instead of hiding those assumptions inside a ranking model.
This does not make Instagram’s algorithm simple. It does make one layer of it more legible. For users, that difference matters because a label can be challenged. A hidden signal cannot.
A Small Example With Real Consequences
Imagine someone searches for wedding venues for two weeks because a friend asked for help. Instagram may start showing bridal fashion, event decor, honeymoon destinations, and venue tours. Under the older model, the user could try to escape that by ignoring posts, tapping “not interested,” or waiting for the system to relearn.
With Your Algorithm on Feed, that person could remove wedding-related topics directly if Instagram has surfaced them as interests. The change is small at the interface level, but it shortens the distance between “Instagram thinks I want this” and “I do not want this in my feed anymore.”
For creators and marketers, the same change has another side. If users can downrank or remove broad topics, weakly targeted content may lose some accidental reach. Content that depends on the algorithm mistaking brief curiosity for durable interest could have a shorter shelf life.
The Bigger Question Is Agency
Instagram’s announcement lands in a period when major social platforms are under pressure to explain how recommendation systems shape attention. Users often like personalization when it works, but resent the feeling that an app is steering them through assumptions they never approved.
Your Algorithm is Instagram’s attempt to make that relationship less one-sided. It acknowledges a basic frustration with modern feeds: people train the system constantly, but mostly through behavioral traces rather than direct choices.
The feature also gives Instagram a practical advantage. If users correct their interests manually, Instagram may get cleaner preference signals than it can infer from passive behavior alone. A person lingering on a post does not always mean they want more of it. A person explicitly adding or removing a topic is a stronger signal.
What To Watch Next
The most important question is how much weight these edits carry. If removing a topic meaningfully changes Feed recommendations, users may treat Your Algorithm as a real control panel. If the same content keeps coming back under adjacent labels, the feature may feel more cosmetic.
There are also open questions around granularity. Topic controls are useful, but many feed problems are not purely topical. A user might like fitness content but dislike body-transformation posts. They might want recipes but not diet culture. They might enjoy political analysis but not outrage clips. Instagram’s planned expansion into people, moods, vibes, and content types may be where the tool becomes more useful.
For now, the practical implication is straightforward:
- Users get a clearer way to correct recommendations without relying only on likes, skips, and watch time.
- Creators may face a feed environment where audience intent is more explicit and less dependent on inferred interest.
- Instagram gets a chance to make algorithmic personalization feel less opaque while collecting stronger preference signals.
The update does not end the debate over algorithmic feeds. It does, however, move Instagram away from a model where users could only react to what the system served them. Feed personalization is becoming something people can edit, at least at the topic level, and that changes the expectations around what control over a social app should look like.