A Court Just Ruled That Addictive Design Is a Product Defect. Meta Owes $567 Million.

Big Tech lawsuit headlines usually follow the same arc: eye-watering damages number, years of appeals, quiet settlement, nothing changes. This one reads differently. A New Mexico court ordered Meta to pay $567 million, and the basis wasn’t what anyone posted on Instagram. It was how Instagram was built.

The Plaintiffs Sued the Product, Not the Posts

That distinction is doing an enormous amount of work.

For roughly three decades, American platforms have stood behind Section 230 of the Communications Decency Act — the provision that says a platform isn’t the publisher of what its users post. Someone uploads something harmful, someone gets hurt, and Meta says: we just provided the space. That argument has won a lot of cases. It’s the closest thing the industry has to a legal force field.

So the plaintiffs stopped attacking the posts and went after the product itself. Infinite scroll. Autoplay. Push notifications. Streaks and other consecutive-use rewards. Recommendation algorithms that run on a 14-year-old’s feed exactly the way they run on an adult’s. The claim: these features were engineered to exploit adolescent brain development and maximize time-on-app.

Framed that way, Section 230 doesn’t have much to grip. Nobody is asking a court to hold Meta responsible for a stranger’s words. They’re asking whether a feature Meta designed and shipped is defective. It’s the same logic that says a carmaker can’t escape liability by pointing out that the driver was the one steering — not if the brakes were bad.

Products Liability Just Crossed Into Software

The courtroom argument ran along three familiar tracks, all borrowed from a body of law built for physical goods.

First: was the product defective? Not defective in the sense of a cracked part — defective by design. The standard asks whether a safer alternative design existed and the company declined to adopt it. Time limits on by default for teen accounts. Overnight notifications off. Recommendation intensity dialed down. All of these were technically achievable years ago. That they were achievable and unshipped is precisely the point.

Second: did the company know? Internal documents did the heavy lifting here. Meta’s own research had flagged mental-health deterioration among specific user cohorts, and engagement metrics kept winning anyway. Not knowing is negligence. Knowing and proceeding is a different category of exposure — and it’s the reason this award cleared half a billion rather than landing in the tens of millions.

Third: were users adequately warned? Cigarettes carry warnings. Prescription drugs carry warnings. Social apps carry a terms-of-service page nobody reads.

Why New Mexico, of All Places

Because the state attorney general brought the case.

When an individual sues, causation is brutal. Meta’s lawyers only need to ask whether this particular teenager’s depression is really traceable to this particular app, and juries reasonably hesitate. When a state sues, the frame shifts to population-level public health, and the question becomes aggregate harm rather than one person’s medical history. That is exactly the path tobacco litigation walked in the 1990s, and opioid litigation walked in the 2010s. Both ended with defendants writing checks with a lot of zeros.

More than 40 states have similar actions filed or in preparation. Once New Mexico’s reasoning starts getting cited in those courtrooms, Meta isn’t defending a case. It’s defending a front.

The Industry Is Already Doing the Math

Meta will appeal. Obviously. And $567 million is under one percent of a single quarter’s revenue — the money isn’t the injury. The precedent is.

You can see the hedging in the product roadmap already. Teen accounts private by default. Overnight notification blocking. Screen-time nudges. These shipped over the past few years partly to get ahead of regulators, and partly so that a lawyer could someday stand in front of a jury and say the company tried.

The underlying contradiction hasn’t moved, though. In an advertising business, time-on-app is revenue. Making the product less compulsive means making less money. Whether that tension can survive self-regulation is the question that decides the next several years — and the honest answer is that no ad-funded company has ever voluntarily solved it.

This Doesn’t Stop at the US Border

Korea has been circling youth-protection rules for social platforms for a while. The country repealed its gaming shutdown law — a decade-old rule that blocked under-16s from online games after midnight — only recently, and the same regulatory energy is now looking for a new target. The UK’s Online Safety Act and the EU’s Digital Services Act are pushing from a different direction, with age-assurance mandates and design-based duties of care already on the books.

The mechanisms differ. The US applies pressure through damages awards and discovery. Europe and Asia legislate first and litigate later. But the destination is the same: the design of the service is now the regulated surface, not just the content flowing through it. Anyone shipping infinite scroll and aggressive notification schedules should be running the numbers on how long that stays a free choice.

The verdict is fresh enough that the discourse hasn’t fully formed yet. What happens on appeal, and how fast the reasoning spreads to the other state cases, will determine whether this is a landmark or a footnote.

The one-line takeaway: a software feature can be found defective the same way a brake line can. The obvious counterargument — that adults are responsible for their own consumption — didn’t move the court, because the plaintiffs weren’t adults.

Think about the apps you opened today. The line between using something and being used by it is genuinely hard to locate from the inside. A New Mexico court just decided that when the user is fifteen, that line can be drawn — and that someone can be made to pay for crossing it.