Instagram Killed Gabby Cusato

Gabby Cusato Was 15. She Was a Runner. Instagram Knew She Was Vulnerable.

Karen Cusato opened her daughter’s phone three days after finding her body. Gabby, 15, had been a cross-country runner — fast, competitive, always chasing a better time. According to the family’s complaint, she’d received her first phone in seventh grade, built up multiple Instagram accounts, some hidden from her parents, and was repeatedly served content about dieting and extreme weight loss. She was in outpatient treatment for anorexia at the time. per family’s complaint

In November 2019, Gabby died by suicide. Her family filed a wrongful-death lawsuit against Meta in 2024. As of this update, legal-tracker reporting indicates the case has since settled — without Meta admitting wrongdoing and without the company committing to platform changes as part of the resolution. probable, per Sokolove Law tracker The Cusatos were never alone in this fight. Thousands of families and school districts are pursuing near-identical claims, and the first juries have now weighed in.

The Machine Didn’t Malfunction. It Performed As Designed.

The recommendations that reportedly followed Gabby’s single nutrition search didn’t appear by accident. Recommendation systems built on variable-reward mechanics — the same principle behind slot machine design — flag a user’s early interest, then test progressively more intense variations of that content, measuring which posts hold attention longest. established

The system doesn’t know a user is 15, or in eating-disorder treatment. It knows engagement. Peer-reviewed research on platform design describes how infinite scroll removes natural stopping points and how notification clustering and algorithmic curation manufacture the same variable reinforcement schedules that make certain products difficult to walk away from. established What court filings in 2026 established wasn’t the mechanism — that part was already documented — but how precisely Meta’s own researchers had traced its effect on identifiable vulnerable groups, and what executives did with that knowledge.

What Meta’s Internal Research Actually Said

In October 2021, former Facebook product manager Frances Haugen testified before Congress and supplied thousands of pages of internal Meta research. The findings were not ambiguous.

That alert was never sent. This is no longer just an activist’s characterization — it’s the finding of a New Mexico jury and, separately, of the California jurors who sat through Mark Zuckerberg’s February 2026 testimony. Two jurors told reporters afterward that his shifting answers on the stand weighed against the company. established

In 2017, internal directives elevated “teen time spent” to a top company metric. In 2019 — the year Gabby died — Meta’s own systems had reportedly flagged roughly 3.5 million profiles pursuing inappropriate interactions with children; teams debated making teen accounts private by default and calculated it would reduce engagement, and did not implement the change until 2024. established

Three Trials, Three Different Legal Theories, One Pattern

These are not the same case. Understanding why matters, because each verdict establishes a different piece of precedent.

Meta’s defense has been consistent across venues: teen mental health has many causes, platforms are used as a coping mechanism for pre-existing difficulty, and no single app can be blamed. In Oakland, Meta’s counsel told the court the company doesn’t dispute that some teens struggle with time management or that people occasionally post negative things — while disputing that this amounts to unlawful design. established

Both companies have appealed the March verdicts. That doesn’t change the settlement math it creates for the roughly 2,000-plus federal cases and 3,000-plus consolidated state cases still pending: a jury saying “liable” once tells every plaintiff’s attorney what juries are willing to believe, and tells Meta’s settlement team what juries are willing to award. probable

Why Section 230 Hasn’t Saved Them — and Why the August Ruling Matters

Starting in November 2023, U.S. District Judge Yvonne Gonzalez Rogers — overseeing the consolidated federal litigation, MDL No. 3047, in Oakland — issued a series of rulings drawing a specific line. Section 230 of the Communications Decency Act shields platforms from liability for what users post. It does not shield a company from liability for how it designed the product itself — infinite scroll, notification clustering, algorithmic amplification. Those are the platform’s own choices, not user content. established

Meta and TikTok tried to get that theory tossed out on appeal before trial. On August 10, 2026, the 9th Circuit refused. Writing for the panel, Judge Jacqueline Nguyen held that Section 230 functions as a defense to liability, not immunity from being sued — meaning the companies can still raise it at trial, but can’t use it to shut down thousands of cases before a jury ever hears them. The ruling didn’t decide whether Section 230 ultimately protects the companies; it decided they don’t get an early exit. established

That distinction — design liability versus content liability, and defense versus immunity — is the legal engine behind everything else in this article. It’s also why the question will resurface for any recommendation engine or personalized feed that optimizes for engagement without disclosing what it optimizes toward, AI systems included.

Inside the Oakland Trial: “How Meta Hooked Kids”

On August 18, 2026, opening arguments began in the largest case yet: a federal trial in Oakland where California, Colorado, Kentucky and New Jersey are leading a 29-state coalition first filed in 2023. California Deputy Attorney General Megan O’Neill told the court the states would show established how Meta designed its platforms so children kept returning, despite internal research documenting traumatic experiences.

Meta’s lead attorney, Paul Schmidt, told the court the company does not dispute that some teens struggle to manage their time on social media or that people post negative content — while rejecting the states’ broader design-liability theory. California Attorney General Rob Bonta has drawn an explicit comparison to 1990s tobacco litigation, telling reporters he believes Meta can redesign its products to be safer established “and still be wildly successful as a business.” The case is expected to run several weeks; the remaining 25 states in the original coalition will go to trial separately, later.

Thousands of Cases. The Math Is No Longer Theoretical.

Plaintiffs’ attorneys in the federal MDL had estimated wrongful-death case values in the high six to low seven figures before any verdict existed. The March and August rulings validate that floor and, in New Mexico’s case, dramatically exceed it. Additional bellwether pools are moving through the pipeline behind the Oakland trial, alongside school-district public-nuisance claims and additional state attorney-general suits — Tennessee has its own state-court trial pending, and Boston recently filed suit on behalf of its public schools. established

Meanwhile, the regulatory track keeps moving on a parallel timeline: California’s Protecting Our Kids from Social Media Addiction Act was largely upheld by the 9th Circuit in 2025; the Surgeon General has called for warning labels on social media comparable to cigarette warnings; and several states are pursuing algorithmic-feed restrictions for minors. None of this guarantees any individual case’s outcome — appeals on the March and August verdicts are already underway — but the design-liability doctrine that makes all of it possible is no longer a novel legal theory. It’s now the operative framework three separate courts have applied.