The Girl, the Jury, and the Algorithm: Inside the Trial That Could Redefine Big Tech

Published: 25 Jul 2026

On February 18, 2026, a man worth roughly $220 billion walked into a courtroom in downtown Los Angeles and sat down across from a jury of twelve strangers.

Mark Zuckerberg had testified before Congress before. He’d sat through televised hearings where grieving parents stood silently behind him, holding photographs of children they’d lost. He’d faced hostile questions from senators and been the subject of documentaries, exposés, and boycotts. None of that was this.

This was the first time in his company’s history that Meta’s own CEO had been called to answer, under oath, before a jury empowered to make him pay — personally putting his platform’s design choices on trial rather than his public image. Cameras were barred from the courtroom. The judge warned that anyone caught wearing AI-enabled smart glasses during his testimony would be held in contempt.

Sitting somewhere in that courtroom was a 20-year-old woman from Chico, California, known publicly only by her initials: K.G.M. Her lawyers called her Kaley. She had downloaded YouTube at age six. Instagram at nine. By her teenage years, according to evidence presented at trial, she was spending up to sixteen hours a day on these platforms.

The question the jury had been assembled to answer sounds almost too simple for a case this large: did the apps do this to her — or did she do this to herself?

Rewind and Setup

To understand why this particular case, involving this particular young woman, ended up carrying so much weight, you have to go back five years — to a moment when the tech industry’s carefully maintained public image cracked for the first time.

In September 2021, a former Meta data engineer named Frances Haugen leaked a trove of internal company documents to the Wall Street Journal. The resulting series, dubbed “the Facebook Files,” revealed something the company had never said publicly: its own researchers had concluded that Instagram made body-image issues worse for roughly one in three teenage girls who already felt bad about their bodies. Internal research linked the app to anxiety and depression in a meaningful share of teen users. A subset of teens, the documents showed, connected their own desire to self-harm directly back to their Instagram use. Haugen testified before the U.S. Senate the following month, telling lawmakers the company knew about these harms and kept its design choices in place anyway.

The revelations landed like a match in dry grass. Lawyers across the country who represented families of children struggling with depression, self-harm, and worse began filing lawsuits against Meta, Google’s YouTube, TikTok, and Snapchat — not over any single video or post, but over the platforms themselves: the endless scroll, the autoplay, the push notifications, the algorithms quietly learning exactly what would keep a thirteen-year-old’s thumb moving for one more hour.

Individually, these lawsuits were unremarkable — one family, one company, one set of allegations. But they kept coming. By early 2026, more than 10,000 individual lawsuits and nearly 800 from school districts had piled up nationwide, consolidated into a sprawling federal multidistrict litigation known as MDL-3047, alongside a parallel set of roughly 1,600 California cases — including over 350 families and 250 school districts — coordinated under a single court proceeding, JCCP 5255.

With that many cases waiting in line, courts needed a way to test the underlying legal theory before scheduling thousands of individual trials. The solution was a bellwether trial: pick one representative case, let it go the distance in front of a real jury, and use the outcome to gauge how the rest might play out.

Out of 1,600 plaintiffs, the first case chosen to test this theory in front of a jury belonged to a young woman from Chico.

The Inciting Event

K.G.M.’s case, formally filed on July 25, 2023, in Los Angeles County Superior Court, described a childhood shaped by four different platforms arriving at four different ages — each one, her legal team argued, engineered by teams of engineers and behavioral scientists to keep her coming back.

According to her attorneys’ opening statements, K.G.M. began using social media around age 10, despite her mother’s attempts to block access with third-party monitoring software. What followed, according to allegations laid out in court filings, included bullying, exposure to content that encouraged self-harm, and a sextortion incident on Instagram — an attempt by a stranger to extort her using explicit material — that her family says went unaddressed by the company for two weeks despite repeated reports from friends and relatives. Her legal team argued that the resulting depression, anxiety, body dysmorphia, and self-harm weren’t a byproduct of ordinary adolescence gone wrong, but the foreseeable result of a product built, tested, and deployed specifically to maximize the time young users spent on it.

Meta and Google saw a fundamentally different story. Their attorneys pointed to K.G.M.’s home life — citing allegations of verbal and physical abuse by her parents — as a more direct explanation for her mental health struggles than any app ever could be. They also raised a harder, more structural defense: that “social media addiction” isn’t a recognized medical diagnosis at all, and that holding a platform legally responsible for how a person chooses to use it sets a dangerous precedent.

Originally, four companies stood accused in K.G.M.’s lawsuit: Meta, Google, Snap, and TikTok. That number didn’t last.

The main question, once the trial actually opened: could a jury be convinced that a piece of software — not a person, not a parent, not a stranger online — was legally responsible for what happened to a child who grew up inside it?

The Investigation and Journey

The case’s slow unraveling of defendants began well before opening statements.

On January 20, 2026, Snapchat’s parent company Snap reached a confidential settlement with K.G.M., roughly a week before trial was set to begin. No amount was disclosed. No liability was admitted. A week later, on January 27, TikTok did the same thing — reportedly settling the very day jury selection was scheduled to start.

Discovery: two of the four original defendants had quietly bought their way out of the biggest test case in the industry’s history. New question: what did Snap and TikTok know, or fear, that made settling — even confidentially, even without admitting fault — look better than facing this particular jury?

That question was left hanging, because settlements are confidential by design. What the public did get, instead, was a trial narrowed down to its two biggest remaining targets: Meta and Google, standing alone in front of twelve Los Angeles jurors, with attorney Mark Lanier leading K.G.M.’s case.

Lanier’s opening statement set the tone immediately, telling jurors that Meta and Google had built, in his words, “machines designed to addict the brains of children” — and that they had done it deliberately, not by accident. Over the following weeks, the plaintiff’s team walked jurors through internal documents echoing Haugen’s 2021 disclosures, alongside expert testimony describing how features like infinite scroll, algorithmic recommendations, autoplay, and push notifications create dopamine-driven feedback loops — the kind that pull a user into what researchers call a near-flow state, where time seems to dissolve and the impulse to keep scrolling overrides deliberate choice.

Discovery: the plaintiff’s team wasn’t just arguing that the content on these platforms was harmful — they were arguing the architecture itself was the defective product, a strategy explicitly designed to sidestep Section 230, the federal law that shields platforms from liability over user-generated content. New question: if the case wasn’t about content, but about design, how would the companies possibly defend engineering choices that millions of ordinary product decisions share — recommendation algorithms, notifications, autoplay — features baked into nearly every app on a modern phone?

Meta’s defense leaned hard into causation. Attorney Paul Schmidt argued that K.G.M. had sought out social media to cope with mental health struggles that existed before she ever opened Instagram — reversing the plaintiff’s causal arrow entirely. The company’s broader legal strategy across the MDL has centered on the claim that as content publishers, platforms cannot be held liable for material posted by third parties, and that reframing “design” as separate from “content” is a legal sleight of hand.

Discovery: both sides were fighting over causation itself — not just what happened to K.G.M., but which came first, her mental health struggles or her app use. New question: how do you prove, to a jury’s satisfaction, which came first inside one particular teenager’s life, a decade after the fact?

Then came February 18 — the day Mark Zuckerberg took the stand.

Under questioning from Lanier, Zuckerberg was pressed on a striking internal detail: that Meta’s own trust-and-safety data had identified roughly 4 million under-13 users on Instagram back in 2018 — accounts that technically violated the platform’s own minimum age policy. Lanier asked directly whether Meta had removed all of them. Zuckerberg acknowledged that it had not, though he said the company had built tools to detect and address underage accounts and continued working to improve those systems. He also acknowledged, more broadly, that children routinely lie about their age to access Instagram, pointing to the company’s longstanding 13-and-over policy as evidence the company had never intended for young children to be there in the first place.

K.G.M.’s team offered a sharper rebuttal: when K.G.M. herself downloaded Instagram at age nine, no meaningful age-verification barrier existed to stop her. The company’s rule, in other words, was real on paper and porous in practice.

Discovery: Meta’s own internal numbers — 4 million known underage users left in place — became some of the most damaging evidence in the entire trial, delivered not by a leaked document, but by the CEO himself, under oath. New question: would that admission be enough on its own — or would the jury still need to be convinced that platform design, and not a difficult home life, was the real cause of what happened to Kaley?

The trial ran for weeks. Jurors heard from behavioral experts on both sides, reviewed internal company research, and weighed testimony about K.G.M.’s life both online and off. Closing statements came on March 12. Deliberations began the next day, March 13 — and did not go smoothly.

Discovery: partway through deliberations, jurors signaled they might be deadlocked on at least one of the two remaining defendants, prompting Judge Carolyn Kuhl to warn that a failure to reach consensus could force a partial retrial. New question: after weeks of Zuckerberg’s testimony, internal documents, and dueling experts, could this landmark case actually end not with a clean verdict, but with a hung jury on one company and a decision on the other?

The Twist

For nine days and roughly forty-three hours of deliberation — an unusually long stretch for a civil trial — the jury wrestled with a case that, on its surface, looked like it should break cleanly one way or the other. Either the platforms were defective products that caused foreseeable harm, or a young woman’s difficult adolescence had causes that had nothing to do with an app.

The twist wasn’t that the jury sided entirely with one story over the other. It’s that they didn’t.

On March 25, 2026, the jury returned its verdict — and it wasn’t a simple win or loss for anyone. Jurors found both Meta and Google negligent, but they didn’t treat the two companies as equally responsible. The damages broke down with striking specificity: $3 million in compensatory damages, split 70 percent against Meta and 30 percent against Google. On top of that, $3 million in punitive damages — $2.1 million assigned to Meta, $900,000 to Google — arriving the very same day as the compensatory verdict.

In other words, the jury didn’t just decide the platforms were responsible. They quietly decided Meta’s Instagram had done more damage than Google’s YouTube — a distinction the public verdict form doesn’t explain, but one that matches the trial’s own evidentiary center of gravity: internal Meta research on teen body image, the sextortion allegations tied specifically to Instagram, and Zuckerberg’s own admissions about underage users all pointed more heavily at one company than the other.

It was also the twist buried inside the earlier settlements. Snap and TikTok had opted out before the jury ever heard a word of testimony — and given how the verdict ultimately landed, with real financial exposure and a genuine liability finding, their decision to settle confidentially rather than risk this jury now reads less like caution and more like foresight.

The Climax

The headline verdict was simple enough to fit in a single sentence: for the first time in the history of the social media industry, a jury had found major tech companies liable — not for content posted by users, but for the deliberate design of their own products, in a case brought by a young person who says those products changed the trajectory of her life.

Six million dollars, spread across two companies, is a modest sum for corporations of Meta and Google’s scale. That was never really the point. The verdict was never primarily about the dollar figure — it was about whether platform architecture itself, the invisible mechanics behind the feed a person scrolls through every day, could be treated by American law the same way a defective car part or a dangerous toy might be: a product, built by a company, capable of causing foreseeable harm regardless of what any individual user chose to post or watch.

For the roughly 1,600 plaintiffs waiting behind K.G.M. in California’s coordinated proceedings — and the nearly 2,900 cases stacked up in the federal MDL — this verdict became the first real data point anyone had. It told plaintiffs’ attorneys nationwide that a jury could, in fact, be persuaded to separate “design” from “content” and hold a platform accountable on those terms. It told the tech industry that its Section 230 shield, long treated as close to bulletproof, has a gap in it — one that plaintiffs’ lawyers are now actively organizing their cases to fit through.

Both companies have stated they intend to appeal.

The Aftermath

The verdict against Meta and Google landed just weeks after a separate, far larger ruling in New Mexico, where Meta was ordered to pay $375 million over allegations it enabled child exploitation and failed to protect minors — a reminder that K.G.M.’s case, as significant as it was, is only one front in a rapidly expanding legal war being fought across multiple states and courtrooms simultaneously.

Inside the industry, the implications are already being treated as more than theoretical. Legal analysts and industry newsletters have described the case as marking the collapse of what one commentary called the “engagement-at-all-costs” model that has quietly underpinned social media’s business logic for two decades — the assumption that platforms could be built, tested, and optimized purely to maximize time spent, without ever being held legally responsible the way a manufacturer might be for a faulty product.

What remains genuinely unresolved is significant. Both Meta and Google are appealing, meaning the $6 million verdict is not yet final, and an appellate court could still narrow or overturn the jury’s findings on legal grounds that have nothing to do with the underlying facts. The verdict also doesn’t establish a binding nationwide precedent — as a state-court jury decision, it directly affects only this case, though its influence on settlement negotiations and litigation strategy across the remaining thousands of pending cases is already visible. And at the center of the legal and scientific debate that shaped this trial, the question Meta’s own attorneys raised in court — whether “social media addiction” constitutes a recognized clinical condition at all — remains genuinely contested among researchers and hasn’t been settled by a single jury’s verdict, however landmark.

For K.G.M. herself, now in her twenties, the trial closes one very public chapter of a story that began when she was a child too young to understand what the app on her mother’s old phone was doing to her attention, her self-image, and her sense of the world. The jury’s verdict didn’t undo any of that. What it did, for the first time, was put a number and a legal finding behind the claim that the platforms themselves bore some of the responsibility — a claim thousands of other families are now watching closely, waiting to see if their own cases will get the same chance in front of a jury.

Sources: Wikipedia summary of K.G.M. v. Meta et al.; NPR; Reuters (via Yahoo Finance and AOL); Fox Business; The Independent (via AOL); FOX/Los Angeles court coverage; The Hill; the MDL Update tracking service (MDL-3047); the University of San Francisco Center for Law, Tech, and Social Good; The Race to the Bottom legal blog; and additional legal-industry reporting on the K.G.M. v. Meta Platforms, Inc. verdict.

A note on this story: it involves themes of self-harm, sextortion, and suicidal ideation among minors. If you or someone you know is struggling, the 988 Suicide & Crisis Lifeline is available by call or text at 988, or online at 988lifeline.org.

The Girl, the Jury, and the Algorithm: Inside the Trial That Could Redefine Big Tech

For twenty years, social media companies operated on a simple legal assumption: platforms aren’t liable for what happens to the people using them. In the spring of 2026, a Los Angeles jury spent forty-three hours deciding whether that assumption was wrong.

Rewind and Setup

The Inciting Event

The Investigation and Journey

The Twist

The Climax

The Aftermath