The Verdict That Broke the Mold
It wasn’t the amount that shocked anyone. Six million dollars? For Meta and Google? A rounding error. But the jury’s verdict—on March 25, 2026—wasn’t about money. It was about accountability.
Kaley, a 20-year-old from Chico, California, had spent her adolescence scrolling. YouTube at six. Instagram at eleven. By fifteen, she was skipping meals to check likes. By seventeen, she’d stopped talking to her family. Her therapist never wrote "social media" on the chart. But Kaley did—on her bathroom mirror, in shaky red lipstick, after another sleepless night.
The jury didn’t need a diagnosis. They needed to know: Did the platforms engineer this? The answer was yes. Not because of what users posted, but because of what the apps did.
Infinite scroll. Autoplay. Notifications that pinged like a heartbeat. Beauty filters that turned Kaley’s face into a stranger she was desperate to become. Meta’s own internal documents showed executives calling tweens "the next growth frontier." Eleven-year-olds were four times more likely to return to Instagram than to any competitor—even though they were supposed to be banned.
This wasn’t a failure of moderation. It was a feature.
And for the first time in history, a jury agreed: digital platforms can be defective products. Not because they host harmful content, but because their architecture is the harm.
The verdict wasn’t just legal. It was cultural. It said: We’re done blaming kids for being weak. We’re done pretending addiction is a personal failing when the system was built to exploit the most vulnerable parts of a developing brain.
The same jury, by the way, didn’t find Meta guilty of causing Kaley’s depression. They found it a "substantial contributing factor." That’s all it took.
The dam cracked that day. And now, 2,200 cases are rushing through the breach.
How Lawyers Outsmarted Section 230
For decades, tech companies had a legal shield. Section 230 of the Communications Decency Act said: You’re not liable for what users say. So if a teenager was bullied on Instagram? That was the bully’s fault. If a child saw self-harm content on YouTube? That was the uploader’s problem.
The law was clear. The loophole was wide.
But the lawyers for Kaley didn’t try to sue over content.
They sued over code.
"It’s not what’s on the screen," lead attorney Mark Lanier told the jury, holding up a jar of M&Ms. "It’s how the screen keeps you looking."
They didn’t argue that YouTube was a bad platform. They argued it was a badly designed one. Infinite scroll? That’s not a feature—it’s a trap. Autoplay? That’s not convenience—it’s compulsion. Notifications? That’s not a reminder—it’s a Pavlovian bell.
They called it a "digital casino." Not because it had gambling, but because it was built on the same principles: variable rewards, unpredictable feedback loops, and a complete absence of exit ramps.
And here’s the genius: Section 230 protects platforms from third-party content. It doesn’t protect them from their own design choices.
The court agreed. In a landmark 2025 ruling, judges declared that digital platforms could be held liable under product liability law if their functionality—specifically, their addictive architecture—was intentionally engineered to exploit psychological vulnerabilities.
This wasn’t about censorship. It was about construction.
Meta’s defense? "We didn’t force anyone to use it."
But the jury saw the internal emails. Zuckerberg’s memo: "We must bring them in as tweens." The engineering team’s metrics: "Time-on-app" as the primary KPI. The design team’s A/B tests: which filter made users spend 27% longer?
The law didn’t change.
The interpretation did.
And now, every time a platform adds a new "engagement" feature—streaks, rewards, algorithmic feeds—it’s not just a product update. It’s a legal risk.
They didn’t break Section 230.
They bypassed it.
And it’s working.
The Wave That’s Coming
The KGM verdict? That was just the first wave.
A day before it, in New Mexico, another jury ordered Meta to pay $375 million—not for depression, but for failing to protect children from predators on Instagram and Facebook. The jury didn’t just find negligence. They found deception. Meta had told parents the platform was safe. Internal data showed they knew it wasn’t.
This wasn’t a civil case. It was a public nuisance claim.
And now, the state of New Mexico is demanding structural changes. Not fines. Not apologies. Actual redesigns. Mandatory safety toggles. Age verification that works. A ban on algorithmic recommendations for users under 18.
This is the new frontier.
The lawsuits aren’t just about compensation anymore. They’re about control.
And the pattern is clear: bellwether cases are setting the tone. The first tobacco verdicts didn’t bankrupt Philip Morris. But they forced them to stop advertising to kids. The first auto safety verdicts didn’t make GM fold. But they made seatbelts mandatory.
Same playbook.
Snapchat and TikTok settled before trial. Why? Because they saw the handwriting on the wall. They knew the jury wouldn’t care about their "community guidelines." They’d care about the algorithm.
The MDL No. 3047—2,200+ cases consolidated under one judge—is now the largest tech liability action in U.S. history. And the plaintiffs’ lawyers aren’t just targeting Meta and Google anymore.
They’re coming for Apple. For Microsoft. For the AI chatbots now being pushed into classrooms.
Because if you can be held liable for designing an app to exploit adolescent dopamine pathways… what happens when you design an AI tutor that rewards students for spending more time on it?
The question isn’t if the next wave will come.
It’s whether we’ll be ready when it hits.
AI on Campus: The Next Trial
Here’s the uncomfortable truth: colleges are rushing into AI the same way they rushed into social media.
No safety studies. No consent forms. No psychological impact assessments. Just "this tool will help you write better." And suddenly, every student has an AI writing coach, an AI study buddy, an AI career advisor—all trained on billions of human interactions, all optimized for engagement.
At Florida International University, Professor Otis Kopp watches students stare at screens, not books. "We’re integrating AI without any psychological drafting of safety protocols," he says. "We’re assuming the brain can handle it. It can’t."
Kopp isn’t anti-tech. He’s pro-awareness.
He’s seen students who used AI to write their papers—then couldn’t explain them in class. Who relied on AI for emotional support, then forgot how to talk to a friend. Who trusted an algorithm’s judgment about their career path—only to realize the AI was trained on data from people who never took a risk.
And then there’s Jessica Kizorek, who says the goal isn’t to ban AI. It’s to build resilience.
"We’re training students for a future we can’t predict," she says. "The goal isn’t perfect answers. It’s the confidence to navigate uncertainty responsibly."
But here’s the twist: the same legal logic that’s being used against social media is already being applied to AI.
In 2025, a Florida court ruled in Garcia v. Character Technologies that an AI chatbot could be considered a "product" under product liability law. The plaintiff claimed the AI reinforced harmful self-perceptions after she asked it to help her feel better. The AI kept feeding her affirmations that were… factually inaccurate. And emotionally addictive.
The case settled. But the precedent was set.
AI isn’t just a tool.
It’s a system. And systems have design.
If you can be sued for making a social media feed addictive, you can be sued for making an AI tutor addictive. If you can be sued for hiding the fact that your platform harms teens, you can be sued for hiding the fact that your AI model is biased.
The courtroom is coming to campus.
And the students? They’re not just users.
They’re the next generation of lawyers, clinicians, and engineers.
They’re the ones who’ll have to answer for what we built.
The Responsibility Question No One Wants to Answer
We keep saying: "It’s the user’s responsibility."
But what if the user is 11?
What if the system was designed to exploit the part of the brain that hasn’t finished developing?
What if the company’s entire business model depends on keeping you hooked?
That’s the paradox.
When corporations are financially driven to maximize engagement, the concept of self-regulation becomes a paradox.
We don’t blame a child for eating candy when the entire store is built to make them crave it. We don’t blame a smoker for lighting up when the packaging is designed to look cool and the ads run during cartoons.
So why do we blame a teenager for scrolling?
The science is clear: adolescent brains are wired for reward. They’re not wired for restraint. Social media didn’t invent addiction. It just found the perfect target.
And now AI is doing the same.
Clinicians are seeing it firsthand. Patients who can’t focus for more than 90 seconds. Who feel anxious if they’re not getting notifications. Who compare their lives to AI-generated idealized versions of others.
It’s not about willpower.
It’s about architecture.
And if we keep pretending this is a personal failure, we’re not just ignoring the truth.
We’re enabling it.
The next decade won’t be decided by how fast we innovate.
It’ll be decided by who we hold responsible when innovation harms.
Will we let companies design systems that exploit children’s brains—and then blame the children for being too weak to resist?
Or will we finally say: Some designs shouldn’t exist.
Because some harms aren’t accidents.
They’re business models.