Court filings in Meta's ongoing lawsuit reveal the company's own research shows 19% of teen users reported seeing unwanted explicit content on Instagram. That's nearly 1 in 5 teenagers. The data was kept internal while Meta publicly claimed its safety systems were working, raising questions about what else the company knows about harm to minors on its platforms.

This is the Frances Haugen playbook repeating itself. Meta has the data showing their platforms harm kids. They choose not to act on it because fixing it would hurt engagement metrics.

What the Court Filings Show

According to The Globe and Mail's reporting, the internal research data emerged in court documents related to lawsuits against Meta over its handling of teen safety. The company's own studies found that a substantial percentage of teen users were exposed to explicit content they didn't request.

This wasn't an external report by critics or activists. This was Meta's internal research team quantifying the problem. They knew. They had the numbers. And they kept building the same systems anyway.

The Engagement Problem

Here's the uncomfortable truth about social media platforms: they're optimized for engagement, not wellbeing. Instagram's algorithm is designed to keep users scrolling. It learns what content keeps each user's attention and serves more of it.

If explicit content drives engagement - even unwanted explicit content that makes users uncomfortable but keeps them on the platform - the algorithm has an incentive to surface it. The system doesn't distinguish between "engaged because interested" and "engaged because disturbed." It just measures time spent.

Meta has spent millions on PR about teen safety features - parental controls, content filtering, age verification, reporting tools. But those are bolt-on features trying to fix a system fundamentally designed to maximize engagement. It's like adding seatbelts to a car that accelerates toward walls.

The Pattern

When Haugen leaked internal Meta documents in 2021, they showed the company knew Instagram was harmful to teen mental health. Meta's response was to question the methodology, emphasize their commitment to safety, and continue largely unchanged.

Now we're seeing the same pattern. Internal research shows significant harm to minors. External pressure forces disclosure. Meta expresses concern and points to safety initiatives. The fundamental incentive structure - maximize engagement at all costs - remains untouched.

Why Algorithms Make It Worse

In the pre-algorithmic era of social media, you saw content from people you followed, chronologically. If someone sent you explicit content, that was a deliberate action by an individual. You could block them.

Algorithmic feeds change the dynamic. Now the platform itself is choosing what you see based on predicted engagement. Instagram isn't a neutral conduit - it's an active participant in what content reaches users. When 19% of teens see unwanted explicit content, that's not just bad actors on the platform. That's the algorithm's choices about what to surface.

What Meta Could Do

The fixes aren't technically impossible. Instagram could deprioritize explicit content in recommendations to teens. They could use their sophisticated computer vision systems to detect and filter such content more aggressively. They could reweight engagement metrics to penalize content that keeps users on the platform but triggers negative reactions.

They don't, because those changes would reduce engagement. Teen users might spend less time on Instagram if the algorithm wasn't constantly serving them controversial or disturbing content. That would hurt revenue.

So Meta keeps the algorithm as is, adds some parental control features that most teens immediately circumvent, and waits for the next scandal.

The Broader Question

This isn't just about Meta. Every major social platform faces the same tension between user wellbeing and engagement optimization. But Meta is unique in how much internal research they've conducted about harm to minors, and how consistently they've chosen not to act on it.

The company knows what their platforms do to kids. They have the data. They're choosing to keep building systems that maximize engagement even when that means exposing teenagers to content that disturbs them. That's not an accident of technology - it's a business decision.