A judge in New Mexico has done something no court had done before: he treated a social network the way the law treats a factory pumping smoke into the air above a town.
On Thursday, Judge Bryan Biedscheid ordered Meta to pay a further $567m (about £421m) for failing to warn the public about the dangers its platforms posed to children. Added to the $375m already imposed earlier in the same case, the state of New Mexico has now extracted $942m from the owner of Facebook, Instagram, WhatsApp and Threads. It is the largest child-safety penalty the company has ever faced.
The number is striking. The legal reasoning is more consequential.
Biedscheid described Meta as a "public nuisance" - a centuries-old legal category normally reserved for things like contaminated water, blocked highways and industrial pollution. In his ruling he compared the company to a factory, with advertising and content as its product and "the psychological harm and sexual exploitation of children to be the pollution that must be abated".
What the court actually decided
The case began with a lawsuit brought by New Mexico state attorneys in 2023. Its central argument was not that Meta hosted bad content by accident, but that the architecture of its platforms endangered children: exposing them to sexually explicit material and putting them within reach of adults who sought to exploit them.
The trial ran in two phases.
In the first phase, the court found Meta had repeatedly violated New Mexico's Unfair Practices Act, concluding that its recommendation algorithms - the systems that decide what a user sees next - had effectively "steered" young users toward harmful content and harmful contacts. That phase produced the $375m figure and made New Mexico the first US state to successfully sue Meta over child safety.
In the second phase, decided this week, the judge moved from consumer-protection law to nuisance law. He found the harms were no longer confined to individual users or even to Meta's own apps. They "migrate to the internet as a whole and, perhaps most concerning, to the real world", he wrote, creating "a common, societal burden on and harm to the affected children and their families" - and on schools, hospitals and law enforcement that absorb the consequences.
| Phase | Finding | Amount |
|---|---|---|
| Phase one | Violations of New Mexico's Unfair Practices Act; algorithms steered minors to harm | $375m |
| Phase two | Harms rise to the level of a public nuisance | $567m |
| Total | $942m |
Where the money goes
Unlike a conventional fine paid into general state revenue, the bulk of this award is earmarked for repair.
- $420m - the vast majority - must fund treatment for harms already caused, through "appropriate clinical or other behavioral health programs and professionals".
- A further tranche funds training and prevention, including for teachers and health professionals learning to recognise and respond to social-media-driven harm in young people.
- Meta must place the money in a dedicated abatement fund aimed at reducing future harm, mirroring how courts handle environmental clean-up.
The orders that may matter more than the money
For a company that reported $61bn in revenue between April and June - up 28% year on year - a nine-figure payment is absorbable. The injunctive terms are harder to shrug off, because they change product behaviour rather than the balance sheet.
Under the order, Meta must:
- Ensure no account belonging to a user under 18 is recommended to an adult.
- Ensure no adult can message an underage user.
- Enact a one-strike policy for adult users who engage in child sexual exploitation.
- Ban push notifications to minors between 22:00 and 07:00, and during the typical school day between 08:00 and 15:00 on weekdays, with weekends exempt.
Read together, those four requirements attack the two mechanics that critics have targeted for years: the discovery loop that surfaces children to strangers, and the notification loop that pulls children back to the app when they should be asleep or in class.
Meta rejects the ruling. "We disagree with the ruling and will appeal," a spokesman said on Thursday, adding: "We work hard to keep people safe on our platforms and have been transparent about the challenges of identifying and removing bad actors and harmful content. We remain confident in our record of protecting teens online and will continue to defend ourselves against claims that misrepresent the facts." The company gave a similar response, and also signalled an appeal, after the earlier $375m verdict.
"A drop in the ocean" - or a template?
Bruce Daisley, a former European vice-president of Twitter (now X), told BBC Radio 4's Today programme that the fine, however large in isolation, was "a drop in the ocean" for a company of Meta's scale. But he framed it as a directional signal: "it's an indication that we are moving to a stage where social media is going to be tackled around the world".
That is the part investors and regulators will study. Meta is currently facing thousands of lawsuits across the United States on similar themes. Earlier this year it lost a landmark case in Los Angeles, where a court found it could be held liable for building addictive platforms - a finding that, like this one, targets design rather than moderation.
Nuisance law is attractive to plaintiffs for a simple reason: it does not require proving that a specific post harmed a specific child. It asks whether an activity has become so widespread in its harm that it burdens the public. That is the framework that produced the opioid settlements, and it is now pointed at a recommendation engine.
The political ripple beyond the United States
Former UK safeguarding minister Jess Phillips told the same Today programme that American court challenges against big tech could make it easier for the UK to act more aggressively without fearing diplomatic friction with Washington. Her argument is essentially about cover: it is harder to characterise domestic regulation as anti-American protectionism when American judges are reaching the same conclusions.
This lands in a year where platform accountability is already contested on several fronts. Apple has mounted a fresh challenge against a UK order seeking access to private user data, and the wider AI and platform security picture has deteriorated, with high-profile breaches at multiple AI companies. Our earlier reporting on how the AI build-out is straining power grids traced a similar pattern: the infrastructure moved faster than the rules around it.
What this changes for parents this week
Nothing in the ruling requires a parent to do anything. But the injunctive terms are a useful checklist, because they describe the settings that a court has now judged necessary:
- Check who can message the account. Restrict direct messages to existing contacts on any teen account.
- Turn off overnight notifications. The court's 22:00-07:00 window is a reasonable household rule regardless of platform.
- Audit recommendations, not just followers. Harm in this case came through suggested content and suggested accounts, not only from people a child chose to follow.
- Treat school hours as offline hours. The 08:00-15:00 restriction exists because notification pressure during class was found to be part of the harm.
Three unresolved questions
Will the appeal succeed? Meta has said it will appeal both phases. Nuisance law applied to a software product is novel, and appellate courts may be less willing than a trial judge to accept the factory analogy. Expect years, not months.
Does the abatement fund work? Environmental abatement funds have a mixed record; money arrives long after harm, and measuring reduction is hard. A $420m behavioural-health fund in a state of roughly two million people is unusually well-capitalised - which makes New Mexico a live experiment other states will watch.
Do other states copy the theory? This is the real stake. A single $942m judgment is a rounding error against Meta's quarterly revenue. Fifty of them, plus injunctive terms that differ state by state, would amount to de facto national product regulation by litigation.
How we got here: a decade of design decisions
To understand why a judge reached for nuisance law, it helps to trace what actually changed inside social platforms over the past decade. The early feed was chronological and finite: you saw what the people you followed posted, and then you ran out. Every layer added since then was designed to remove that stopping point.
Infinite scroll removed the bottom of the page. Algorithmic ranking removed the requirement that you follow someone to see their content. Recommendation surfaces - suggested accounts, suggested reels, suggested groups - turned the product from a directory of your existing relationships into an engine for manufacturing new ones. Push notifications closed the loop by reaching into a pocket to restart a session that had ended.
Each of those decisions is defensible on its own and, for adults, largely uncontroversial. Stacked together and pointed at a thirteen-year-old, they produce the exact pattern the New Mexico court described: a minor who did not seek out harmful content or harmful adults, but was introduced to both by a system optimising for time spent.
That is why the injunctive relief in this ruling matters more than the headline number. The court did not order Meta to moderate harder. It ordered Meta to break specific mechanisms - discovery of minors by adults, messaging from adults to minors, and notifications during sleep and school hours. Those are product specifications, not content policies.
The legal theory, and why companies fear it
Nuisance law developed to solve a problem that individual lawsuits handle badly. If a factory pollutes a river, no single downstream resident can easily prove which illness came from which discharge, and the cost of litigating exceeds any individual recovery. Nuisance law lets a public authority sue on behalf of the shared interest and demand abatement - the clean-up - rather than compensation for named victims.
Applied to software, that framework sidesteps the defences platforms have used successfully for two decades. It does not require identifying a specific illegal post. It does not require proving that a particular child would have been fine but for a particular recommendation. It asks a broader question: has this activity imposed a widespread burden on public health and safety that the operator must now pay to reduce?
The judge answered yes, and located the harm outside the platform boundary. The effects, he wrote, "migrate to the internet as a whole and, perhaps most concerning, to the real world", landing on families, schools, hospitals and law enforcement. That is a description of externalised cost - the same economic concept that underpins environmental regulation.
Meta's appeal will almost certainly attack this framing directly, arguing that a communications service is not a factory, that speech-carrying products cannot be treated as pollution sources, and that federal law preempts state attempts to regulate platform design. Those are serious arguments. They have not yet been tested against this fact pattern at appellate level.
What the numbers say about deterrence
Set the award against Meta''s own disclosures and the deterrence question answers itself for now. Quarterly revenue of $61bn for April to June, growing 28% year on year, means the total New Mexico judgment of $942m represents roughly a week and a half of revenue - and far less of annual profit.
That arithmetic is why the litigation strategy is spreading rather than concluding. Thousands of similar suits are pending across the United States. Meta already lost a landmark Los Angeles case this year on the theory that it could be held liable for building addictive platforms. If a dozen states each obtain both damages and bespoke injunctive terms, the cumulative effect is not a fine - it is a patchwork of legally mandated product requirements that a single global app has to satisfy simultaneously.
Companies usually respond to that pattern in one of two ways: they lobby for a single federal standard that preempts the patchwork, or they build the strictest requirement into the global product because maintaining fifty variants is more expensive than compliance. Both outcomes look like regulation. Neither requires Congress to pass a law.
What to watch over the next three months
- The notice of appeal and whether Meta seeks a stay of the injunctive terms. If the product requirements take effect during appeal, users will see the change before the law is settled.
- Copycat filings. State attorneys general coordinate closely; a successful nuisance theory travels fast.
- Product telemetry. Independent researchers will look for evidence that overnight notifications to teen accounts actually stop, in New Mexico and beyond.
- The abatement fund''s governance. Who administers $420m of behavioural-health spending, and how outcomes are measured, will determine whether this becomes a model or a cautionary tale.
FAQ
How much has Meta been ordered to pay in New Mexico? $942m in total: $375m in the first phase of the trial and a further $567m in the second phase decided this week.
What does "public nuisance" mean here? It is a legal finding that an activity harms a shared public interest so broadly that it must be abated. The judge compared the psychological harm and sexual exploitation of children to "noxious pollution produced by the factory".
Is this the first ruling of its kind? It appears to be the first time a social media company has been formally deemed a public nuisance, and it is the largest child-safety fine Meta has received.
Is Meta paying immediately? No. Meta says it disagrees with the ruling and will appeal, as it did with the earlier $375m verdict.
What product changes are required? No recommending under-18 accounts to adults, no adult-to-minor messaging, a one-strike policy for adults engaged in child sexual exploitation, and bans on push notifications to minors overnight and during school hours.
Sources and further reading
- BBC News: Meta fined $567m in largest child safety ruling against social media giant
- New Mexico Department of Justice: Office of the Attorney General
- Reuters technology coverage: reuters.com/technology
More from Protunez: Technology · Business & Economy · Health & Science
