Instagram Expands Algorithm Customization Tools in Latest Testing Phase

The algorithm becomes less of a black box and more of a tool users can configure.
Instagram's customization tests could fundamentally reshape how platforms operate and how they're regulated.

For years, social media users have scrolled through feeds shaped by invisible hands, accepting algorithmic decisions they could neither see nor influence. Now Instagram, under the direction of Adam Mosseri, is testing tools that would allow users to actively shape their own algorithmic experience — a quiet but significant shift in the relationship between platform and person. The move arrives alongside similar experiments at Threads and TikTok, suggesting that mounting regulatory scrutiny and user frustration have reached a threshold the industry can no longer ignore. Whether this represents genuine democratization of attention or a carefully managed illusion of control is the deeper question these tests will eventually have to answer.

  • Users have spent years watching content appear in their feeds without understanding why, and that frustration has quietly accumulated into a demand the platforms can no longer dismiss.
  • Regulators across multiple countries are now scrutinizing algorithmic mechanics, creating real legal and reputational pressure on Instagram, TikTok, and Threads to demonstrate transparency.
  • Adam Mosseri is actively showcasing early concepts for feed customization tools, signaling that Instagram is moving — however cautiously — toward giving users a hand on the wheel.
  • The features remain in concept phase, unavailable to the general public, and their actual design will determine whether they deliver genuine agency or merely the appearance of it.
  • If these tools prove meaningful, they could shift regulatory conversations, redefine how platforms measure engagement, and fundamentally alter the contract between algorithm and user.

Instagram is testing tools that would let users actively shape their own algorithmic feeds — a departure from the opaque, system-driven model that has defined the platform for years. Adam Mosseri has been demonstrating early concepts for how this might work, though the features are not yet available to the general public. The direction, however, is clear.

The move is not happening in isolation. Threads and TikTok are running parallel experiments, and the timing reflects a broader industry reckoning. Regulators in multiple countries have begun examining how algorithms operate, what they amplify, and whether platforms are being honest about the mechanics of content distribution. Giving users more control is one response to that pressure — and it also shifts some responsibility onto users themselves.

There is a small irony embedded in the moment: the specifics of how these customization tools will function remain somewhat unclear, even as the stated goal is greater transparency. And the deeper question lingers — whether a feature is genuinely useful depends entirely on how it is built. A tool buried in settings serves a different purpose than one that is prominent and intuitive. A feature that makes marginal adjustments while the core algorithm runs unchanged is not the same as one that meaningfully alters what a user sees.

What these tests will ultimately reveal is whether algorithmic transparency and user agency can coexist with the engagement-driven business model social platforms have built their fortunes on. That question remains open. For now, Instagram is watching, listening, and testing — and the answers, when they come, will matter well beyond the platform itself.

Instagram is testing new ways to let users shape what appears in their feeds, moving beyond the black box of algorithmic recommendation that has defined the platform for years. Adam Mosseri, who leads Instagram, has been demonstrating early concepts for how this might work—tools that would let people actively steer their own algorithmic experience rather than simply accepting what the system decides to show them.

The shift reflects a broader reckoning across social platforms. Users have long complained about opacity: they see content, but they don't understand why. The algorithm decides, and that's that. Instagram's testing suggests the company is listening to those complaints, or at least responding to the pressure mounting around them. Threads and TikTok are running similar experiments, each trying to give users more granular control over their feeds. It's not coincidental timing. Regulators in multiple countries have begun scrutinizing how algorithms work, what they optimize for, and whether platforms are being transparent enough about the mechanics of content distribution.

What Mosseri is showing off remains in the concept phase—these are not yet features available to the general user base. But the direction is clear: Instagram wants to move toward a model where customization is possible, where users can adjust their experience without having to abandon the platform entirely or resort to workarounds. The specifics of how this will function remain somewhat opaque, which is itself a small irony given the stated goal of algorithmic transparency.

The timing matters. Social media platforms have faced sustained criticism over the past several years about algorithmic amplification of divisive content, misinformation, and material that harms mental health—particularly among younger users. Giving users more control over what they see is one way to address those concerns, or at least to appear to address them. It also shifts some responsibility onto the user: if you don't like what you're seeing, you can now adjust the settings yourself.

These customization tools could reshape how platforms operate. If users can meaningfully control their feeds, the relationship between platform and user changes. The algorithm becomes less of a black box and more of a tool that users can configure. That has implications for how platforms measure success, how they think about engagement, and what they optimize for. It also has implications for regulation. If platforms can demonstrate that users have genuine control over algorithmic recommendations, that becomes a stronger defense against regulatory pressure.

What remains to be seen is whether these tools will be genuinely useful or merely performative—whether they'll give users real agency or just the appearance of it. The difference matters. A customization feature that is buried in settings and difficult to use serves a different purpose than one that is prominent and intuitive. Similarly, a feature that actually changes what you see is different from one that makes marginal adjustments while the core algorithm continues to operate as before.

For now, Instagram is in the testing phase. Mosseri is showing concepts. The company is watching how users respond, what they want, what they'll actually use. The broader question—whether algorithmic transparency and user control can coexist with the engagement-driven business model that social platforms have built—remains unresolved. These tests may provide some answers.

Adam Mosseri is demonstrating early concepts for how users could actively steer their algorithmic experience
— Instagram leadership
The Hearth Conversation Another angle on the story
Inventor

Why is Instagram doing this now? What changed?

Model

Regulators started asking harder questions about how algorithms work. Users have been complaining for years. Instagram is responding to both at once—it's partly genuine, partly defensive.

Inventor

But if users can customize their feeds, doesn't that hurt engagement? Algorithms are designed to keep people scrolling.

Model

That's the tension. Yes, it could. But if Instagram doesn't offer some control, regulators might force it anyway. Better to do it on your own terms.

Inventor

Are these tools actually useful, or just window dressing?

Model

That's the real question. We won't know until they're live and we see how they work. A buried setting is different from something prominent and intuitive.

Inventor

What happens if other platforms don't follow suit?

Model

They probably will. TikTok and Threads are already testing similar features. It's becoming table stakes—users expect some control now.

Inventor

Does this actually solve the problem of algorithmic harm?

Model

Not entirely. It shifts responsibility to the user. But it's a start, and it changes the conversation about what platforms owe their users.

Contact Us FAQ