Instagram's New 'Your Algorithm' Feature Gives You Control Over Your Feed—Here's What It Actually Does

Instagram has launched a feature called 'Your Algorithm' that lets you see and adjust the topics that shape what you see in Reels and Explore — a significant move toward transparency on a system that, until now, mostly worked invisibly behind the scenes from your perspective.

The feature was announced by Instagram in December 2025. For the first time, it shows you the specific interest categories the platform's recommendation system has linked to your account. You can review those topics and make direct changes — removing interests that don't fit anymore, or boosting ones that do — with the goal of changing the content mix that appears in your feed going forward.

What 'Your Algorithm' Actually Does

'Your Algorithm' is essentially a transparency and control layer sitting on top of Instagram's existing content ranking system. It does not show you the underlying mathematical model, the engagement signals, or how Instagram traces through networks to find content — instead, it gives you a human-readable summary of the interest categories the system has figured out from your behaviour.

This distinction matters. What you see is not the algorithm itself, but a simplified summary of what Instagram's system thinks you're interested in, based on how you've behaved. The actual ranking system — which considers how long you watch videos, how fast you interact, how recent the content is, how much you like the creator, and many other hidden signals — still runs in the background. What changes is how much weight your stated preferences get compared to what the system infers from your behaviour.

In practical terms: if Instagram's system has categorized you as interested in fitness and home improvement, but you've moved on to other things, 'Your Algorithm' lets you fix that mismatch directly. Without this feature, you'd have to skip or hide content repeatedly until the system picks up on the shift on its own.

The feature works on both Reels — Instagram's short-form videos, which are now the main way people discover and engage with content on the platform — and the Explore tab, which is Instagram's broader content discovery page.

The Regulatory and Competitive Context

It would be incomplete to look at 'Your Algorithm' as just a product feature disconnected from the rules that Instagram's parent company, Meta, has to follow. The European Union's Digital Services Act (DSA), which became fully active in 2023 for the largest online platforms, requires that recommendation systems give users at least one option that does not rely on personal profiling. Platforms must also explain what information is driving your personalized recommendations and let you change or turn off those parameters.

Instagram's new feature aligns closely with those DSA requirements. While Meta hasn't called this a compliance move, the timing and what it includes match what the regulation asks for.

At the same time, TikTok — the competitor that pushed Instagram to build Reels in the first place — has offered similar interest management tools for some time. YouTube has long let users mark content as "Not interested" or block entire channels on Shorts and the main feed. Instagram arriving at a named, organized control system in late 2025 means the major short-form video platforms now generally offer this kind of user control, even if the underlying systems work differently.

There is an important caveat worth considering here. Having a user-facing control option does not guarantee the controls actually work or that your feedback gets processed reliably. Platforms have a mixed track record on this. The gap between "you said you don't want this topic" and "the recommendation engine has actually changed to give you less of it" can be surprisingly wide. In practice, hidden signals — like how many times you watch a video, if you rewatch it, or if you share it — tend to override what you explicitly say you want fairly quickly. Whether Instagram's system will actually weight your corrections heavily enough to produce a real, lasting change in what you see is something we'll only learn by watching how it works in the real world, not from a product announcement.

Why This Matters

From a technology design perspective, 'Your Algorithm' is worth understanding as a puzzle about how to balance two conflicting types of information.

The challenge is this: the system can see what you actually engage with (which is noisy but constant and detailed), and it can ask you what you want (which is precise but rare and sparse). Designing a system that lets your stated preferences matter without breaking the recommendation quality that keeps people engaged is genuinely difficult — and there is no standard, agreed-upon way to do it well.

This pattern has appeared before in the tech industry. When algorithmic social feeds replaced RSS readers in the early 2010s, the idea was that computers could figure out what you'd find relevant better than you could by picking your own sources. Often they could, in terms of keeping you engaged — but that same power to predict also meant you had less control. The years that followed brought growing frustration and regulatory pressure, which has now led platforms like Instagram to put some control back in your hands. This is not a criticism of the industry — it is more or less how consumer technology evolves, one step at a time, especially when there is external pressure pushing change forward.

What Comes Next

Instagram hasn't said whether 'Your Algorithm' will expand to give you finer controls — for example, the ability to manually turn up or turn down specific creators, or to prefer certain video lengths or styles within Reels. That would be a logical next step if enough users engage with the current features.

A bigger open question is whether Meta will share data about how 'Your Algorithm' is actually being used. If they published numbers on how many people use the feature, which topics they change most often, and whether those changes actually shift what appears in their feeds, that would give researchers and regulators a clearer view of whether user controls are genuinely working or are just a surface-level fix. For now, the feature is a visible step toward giving you more say in what you see on a platform used by billions of people. Whether that control is real and lasting remains to be seen.