A New Mexico judge has pushed Meta's liability in a landmark child safety case past $900 million. In a separate case, a Delaware court has ruled that Meta's own insurers don't have to pay for its defence. Taken together, the two rulings say something the UK market shouldn't ignore: social media harm is turning into a real liability class, and a largely uninsured one.
On Thursday, Santa Fe judge Bryan Biedscheid ordered Meta to pay an additional $567 million, on top of the $375 million a jury had already awarded in March. That brings the total penalty to $942 million, the largest sum yet awarded against a social media company over harm to children.
The case was brought by New Mexico's attorney general in 2023, who accused Meta of exposing children on Facebook and Instagram to sexual exploitation, solicitation and human trafficking, and of knowingly allowing the platforms to cause harm.
Biedscheid's order leans on an unusual comparison. He described Meta's business as operating like a factory, with the psychological harm and sexual exploitation of children as the "pollution" it produced, and found the company to be a "public nuisance" – a legal theory more often used against polluters, opioid manufacturers and gun makers than tech platforms. Around $420 million of the award is earmarked for treatment programmes for affected children, with the remainder going toward prevention and awareness work.
Meta says it will appeal. "We disagree with the ruling and will appeal," a company spokesperson said, adding that Meta has worked to be transparent about the difficulty of removing harmful content and bad actors from its platforms.
The New Mexico case sits alongside a much larger wave of litigation. More than a thousand US school districts and attorneys general from 43 states have filed similar claims, now consolidated into what's known as the Social Media Litigation in California. A separate Los Angeles jury ordered Meta and Google to pay a combined $6 million to a single 20-year-old claimant in March, over near-identical allegations that the platforms were designed to be addictive.
The figure that should really catch a UK underwriter's eye didn't come out of New Mexico at all. Weeks earlier, a Delaware Superior Court judge ruled that more than 20 insurers, including Hartford and Chubb, have no duty to defend Meta in the underlying Social Media Litigation. The reasoning: standard commercial general liability policies only respond to "occurrences", broadly meaning accidental or unintended events, and the court found that a platform deliberately built to maximise engagement can't credibly argue the resulting harm was an accident.
That ruling only settles the duty to defend, not any eventual payout, and it's likely to be appealed. But its implications reach well beyond one company's legal bill. Insurers are already pointing to it in other platform-harm disputes, and it echoes similar "intentional acts" findings in opioid and firearms litigation. Courts, in other words, seem increasingly willing to treat algorithm-driven engagement design the same way they've treated other products found to cause mass harm.
This isn't a new theme for this market. We've reported before on how rising "nuclear verdicts" and social inflation are putting outsized pressure on the upper layers of liability towers, and more recently on how addictive design lawsuits are putting insurers on watch as a potentially scalable new source of claims. The shift is from content moderation failures, which platforms have long argued they're shielded from, toward product design, where that shield doesn't apply.
It would be easy to file this under "American problem, American law." That would be short-sighted. The Online Safety Act 2023 has moved from statute book to active enforcement, and Ofcom can now fine platforms up to £18 million or 10% of global revenue for breaches. A government consultation on a possible minimum age for social media use and restrictions on "risky" design features closed at the end of May.
Legal advisers have already flagged where this is heading. A panel of liability defence specialists convened by law firm Kennedys concluded this year that the Act is likely to expand D&O exposure for UK platform directors specifically, as regulation moves toward holding leadership personally accountable for governance failures rather than simply fining the corporate entity.
Insurance Business UK has covered the practical fallout of this already. Kenny Carmichael, technology practice leader at CFC Underwriting in London, has argued that the market has been too narrow in how it categorises this risk, treating platform businesses mainly as data risks when the real exposure spans cyber, media, product liability and casualty all at once. Cyber wordings, in particular, tend to be built around clear breach-or-attack scenarios, not around slow-building psychological harm caused by how a product is designed to work.
A few things worth checking on any book with technology, media or platform exposure. General liability and cyber wordings shouldn't be assumed to respond automatically to "harm by design" claims; the Delaware ruling shows insurers will contest coverage wherever harm can be framed as the predictable result of a deliberate product choice rather than a genuine accident. D&O books touching technology or social platforms deserve a fresh look against the direction of UK regulation, not just historical claims experience, given how new and largely untested the Online Safety Act's personal accountability provisions still are. And the exposure isn't limited to the household names – any client running an algorithm-driven, engagement-optimised platform is a plausible defendant in this style of claim, whether or not they've been sued yet.
Both the New Mexico penalty and the Delaware coverage ruling are being challenged, so neither is the last word. But a nine-figure US verdict built on public nuisance law, sitting alongside a coverage ruling that strips a $1.5 trillion company of its own defence costs, points in one clear direction. UK insurers with any exposure to platform businesses have good reason to start asking harder questions now.
