Social media companies lost their bid on Monday to cut short one of the largest product liability fights in the country.
The 9th U.S. Circuit Court of Appeals has just ruled that Meta, Google, TikTok and Snapchat can't yet appeal a lower court's decision forcing them to face a sprawling multidistrict litigation over claims their platforms were engineered to hook young users. A three-judge panel in San Francisco found the companies had moved too early. This kind of appeal is normally reserved for cases that have already been resolved, not ones still in the middle of pretrial fights.
At issue was whether Section 230 of the Communications Decency Act, the 1996 law that generally shields internet platforms from liability for content posted by their users, also gives them the right to challenge lawsuits before trial. The companies argued the statute amounted to outright immunity, entitling them to an immediate appeal. The panel disagreed. Section 230 operates as a defense that can be raised and won at trial, the judges held, not a jurisdictional shield that stops a case from proceeding at all. That distinction matters for scheduling: the underlying litigation, consolidated before U.S. District Judge Yvonne Gonzalez Rogers in Oakland as In re: Social Media Adolescent Addiction/Personal Injury Products Liability Litigation (MDL No. 3047), keeps moving toward trial instead of freezing while an appeals court works through the legal theory.
The same panel also declined to pause a separate trial that opened this week in California, brought by attorneys general from 29 states, alleging Meta illegally collected children's data, built its platforms to maximize compulsive use, and misled the public about the risks. Meta had wanted that trial put on hold pending its appeal. The court said no.
Section 230 is the legal hook, but the number that matters more to underwriters is the scale of what's still moving toward trial. The federal docket alone now covers more than 3,000 lawsuits filed by parents, school districts, municipalities and state governments. All of them press a similar theory: that features like infinite scroll, autoplay and algorithmic recommendation engines were built to keep children engaged, and that the resulting rise in anxiety, depression and self-harm among young users was a foreseeable, even intended, result of product design rather than a side effect of how people chose to use the apps. A parallel state-court proceeding in California is running alongside the federal cases, with roughly 3,300 additional claims of its own.
Framing the harm as a design choice rather than an accident is exactly what has already cost Meta its insurance defense funding in one major venue. As Insurance Business reported in March, a Delaware Superior Court judge sided with Hartford, Chubb and more than a dozen other insurers who argued that Meta's commercial general liability and umbrella policies don't have to fund its defense in the youth-addiction suits, because a deliberately engineered product isn't an "accident" under California law, which governed the dispute. The court's logic was straightforward: strip away the negligence language in the complaints, and what's left is a company that consciously built a product to maximize engagement. Adding a negligence count doesn't turn that into an accident.
That ruling addressed only the duty to defend, not whether insurers ultimately owe indemnification if Meta loses at trial, and it's still subject to appeal. But it lines up with what coverage specialists have been warning for months: that standard general liability wordings were never written with algorithmic engagement design in mind. Rosehana Amin, a partner at Clyde & Co in London who has advised on coverage questions arising from the same underlying Meta litigation, has said insurers will likely argue that intentional-conduct or "expected/intended injury" exclusions apply wherever the pleadings allege deliberate engagement engineering with known harms, while plaintiffs' lawyers counter that the intent behind the design was to boost usage, not to cause injury. Internal documents from platform employees describing what companies knew, and when, will probably decide how that argument plays out, in the Delaware dispute and in whatever coverage fights follow it.
Monday's ruling arrives on top of a string of losses for the platforms. A Los Angeles jury found Meta and Google negligent in March for building products that harmed a young user, awarding $6 million in compensatory damages split roughly 70/30 between the two companies, in a case Insurance Business covered as an early test of how the industry's "Big Tobacco" moment might unfold for social platforms. Punitive damages, which most general liability and umbrella policies exclude on public policy grounds, are still being litigated in a second phase of that case. Any punitive award will land directly on Meta's and Google's own balance sheets, not their insurers'.
New Mexico has produced the biggest numbers so far. A jury there ordered Meta to pay $375 million in March after finding the company misled consumers about platform safety, and a judge added a further $567 million plus mandated youth-safety measures after finding Meta had created a public nuisance, pushing that single case's total close to $1 billion. Both Meta and Google have said they'll appeal their respective losses. Neither company nor a spokesperson for the plaintiffs responded to requests for comment on Monday's ruling, per Reuters.
Public nuisance claims carry their own coverage wrinkle: they often don't require a showing of bodily injury, which puts them at odds with general liability triggers written around physical or psychological harm to a specific person. Legal teams tracking the space, have flagged aggregation language, defense-cost modeling and policy endorsements as areas underwriters need to test now, not after the next verdict lands.
It's tempting for casualty and tech E&O underwriters outside the handful of trillion-dollar platforms in this litigation to treat the exposure as somebody else's. That's getting harder to justify. Insurance Business has previously noted that the theory being tested against Meta, Google, TikTok and Snap, that addictive design itself is the liability trigger and not just the content sitting on top of it, doesn't stop at social networks. Any digital product built around engagement mechanics, from gaming apps to community features bolted onto retail or fitness platforms, could face a version of the same argument. Smaller companies without Meta's legal budget or layered insurance towers may actually be more exposed to an adverse early ruling than the platforms making headlines right now.
Monday's decision doesn't change any of that calculus. It just confirms that the underlying cases, and the legal theories inside them, are going to keep generating trial verdicts, appellate rulings and coverage disputes for years rather than getting cut off early on a procedural technicality. More litigation reaching more juries means more data for underwriters trying to price this risk. It also means more uncertainty sitting on carriers' books while that data accumulates.
