For years, the line from Meta was consistent and practiced: its platforms are tools, and tools don't cause harm — people do. That defense is now being tested in a federal courtroom in California, where 30 state attorneys general have gathered to argue that the company knew exactly what its products were doing to teenagers and kept building them anyway.

The trial, which stems from a coordinated lawsuit filed in 2023, is not a class action brought by individual parents, and it is not a shareholder grievance. It is a direct assault by state governments on the architecture of the most-used social media platforms on earth. The states are seeking damages that could exceed $1 trillion and are demanding structural changes to the way Instagram and Facebook operate — including how the platforms engage minors, deploy algorithmic recommendation systems, and collect behavioral data on users under 18.

The core of the states' case is that Meta violated both federal children's privacy law — specifically the Children's Online Privacy Protection Act — and a web of state consumer protection statutes. But the more combustible allegation is this: that Meta possessed internal research demonstrating measurable psychological harm to young users, particularly adolescent girls, and not only failed to act on that research but actively suppressed it from public view while continuing to optimize its engagement systems toward the same age group.

That internal research is not a rumor. Documents subpoenaed during discovery include communications and studies conducted within Meta's own data science and product teams. Among the findings that have surfaced in pretrial proceedings: company researchers had identified correlations between heavy Instagram use and increased rates of anxiety, depression, and body dysmorphia in teenage girls. The phrase that has already ricocheted through early proceedings — "the young ones are the best ones" — is attributed to internal Meta communications and is being presented by the states as evidence of a deliberate targeting posture, not a passive side effect.

Meta's legal team has pushed back hard on that framing. The company's position is that correlation in internal research does not establish causation, that the platforms provide enormous social value including to young people, and that the states are attempting to hold a technology company liable for the broader mental health crisis among American youth — a crisis with many documented drivers. Meta has also argued that some of the changes the states demand would effectively require the company to redesign core features in ways that conflict with First Amendment protections around speech and editorial discretion.

The First Amendment argument is not trivial, and legal scholars are divided on how far it reaches. But the states have a counter: they are not telling Meta what content to host. They are alleging that specific engineering decisions — infinite scroll, push notifications timed to psychological vulnerability windows, algorithmic amplification of content that provokes social comparison — constitute unfair and deceptive business practices that targeted a protected class of users: children. That is a products liability and consumer protection frame, not a speech frame, and it is deliberately chosen.

What makes this trial structurally different from prior Big Tech legal skirmishes is the breadth of the coalition and the identity of the plaintiffs. When individual families sue, Meta can litigate them into the ground one by one. When 30 state governments move as a unit — including California, New York, Colorado, and Kentucky — the discovery leverage, the political staying power, and the precedent-setting stakes are categorically different. Kentucky's attorney general has compared this moment to the state-led tobacco litigation of the 1990s and the opioid settlement campaigns of the 2010s. That is not an idle analogy: both of those fights produced structural industry changes that no single lawsuit or regulator could have forced alone.

The $1 trillion damages figure has drawn skepticism from legal observers who note that courts rarely award sums at that scale, and that Meta's total market capitalization — itself enormous — is a fraction of that number. But the damages demand may be less important than what the states are asking the court to order prospectively: mandatory changes to how Meta builds and deploys features that touch users under 18. A court order requiring algorithmic transparency, restricting certain notification behaviors, or mandating genuine age verification would reshape the product in ways no self-regulatory pledge ever has.

What is confirmed going into this trial is the following: Meta's internal teams studied the effect of its platforms on young users, found troubling signals, and the company did not disclose those findings to regulators or the public. What remains to be litigated is whether that constitutes illegal conduct — and whether a jury will decide that the company that built the feed also built the harm.