Meta is back in court. This time, the plaintiffs are 28 states, led by California, Colorado, Kentucky, and New Jersey, which allege that the company deliberately designed their products to get young people addicted, and harmed their mental health as a result—complete with a Johnnie Cochran-esque slogan that Meta “hooks” (users), “holds” (them on the platform), “harvests” (their data), and “hides” (from accountability).
It comes on the heels of multiple losses for the social media giant over substantially the same conduct. Meta was fined $567 million and could be ordered to make changes to its platforms after an early-August ruling out of New Mexico. A previous trial in New Mexico this spring cost the company $375 million, and a separate case in Los Angeles found Meta guilty of $4.2 million in damages for addicting a single young user.
But this case would raise those penalties significantly higher, up to $200 billion or by some estimates $1.4 trillion, equivalent to Meta’s entire market value. That would be similar in scope to the 1996 Tobacco Master Settlement Agreement, whereby cigarette companies pledged $206 billion to medical treatment and anti-smoking messaging, while agreeing to alter some of their own marketing. This Meta case could also add structural remedies that at their most aggressive would end Facebook’s brand of social media as we know it. I’m not sure anybody would be particularly upset.
The states’ argument is based on both state-level consumer protection laws and the federal Children’s Online Privacy Protection Act. The plaintiffs have produced evidence that Meta publicly pronounced its platforms to be safe for children, while privately admitting that they weren’t. Former Meta safety engineer Arturo Béjar testified that he had repeatedly warned top executives about Meta algorithms delivering content from sexual predators or containing graphic violence to children, and they had done nothing to stop it.
The states’ argument is based on both state-level consumer protection laws and the federal Children’s Online Privacy Protection Act.
The facts of the case are not really in doubt. It has already been widely reported in numerous venues that Meta knew their product design decisions were harming young people, particularly girls who have developed body dysmorphia or eating disorders from Instagram, and did not change them. That is one of the principal findings of The Wall Street Journal’s “Facebook Files” series, based on extensive internal documents, published five years ago.
There are two additional important pieces of legal background. One is a 2024 Supreme Court decision, Moody v. NetChoice, LLC, which held that content selection algorithms are protected by the First Amendment. Any proposal to regulate them therefore faces “strict scrutiny” from the Court (at least under this majority), a test which is almost never passed. Second is Section 230 of the Communications Decency Act, which immunizes internet companies from most legal liability over third-party content that they host. If someone defames you in the YouTube comments, for instance, you can sue the person who posted it, but not YouTube.
The states, however, have crafted their lawsuit to get around Section 230. Rather than try to hold Meta accountable for third-party content, as Stanford Law professor Nora Freeman Engstrom explains, “they’re arguing that Meta itself possessed information, made representations inconsistent with that information, misled users, parents, and regulators, and, ultimately, inflicted harm.”
How this kind of lawsuit eventually shakes out will likely depend on whether the courts buy this framing. As noted, three judges have already bought the basic contours of the argument. But those were acceptable losses for Meta; this case is much, much bigger. If Meta wins this one, it’s easy to imagine the creation of a legal catch-22 for social media critics—state governments can’t regulate content algorithms because that’s protected speech, but users harmed by Meta’s atrocious business practices can’t sue either because of Section 230.
On the other hand, it stands to reason that if content algorithms are speech, then social media companies should be legally considered publishers, and hence not enjoy the protections of Section 230 at all. Surely a website hosting provider who rents out server space is doing something quite different than a social media company that actively picks and chooses what content to show each individual—and particularly when the algorithm is a proprietary black box. When we at the Prospect pick an article to publish in our magazine, we are liable for potential defamation claims in addition to the author. It doesn’t seem like too much to hold Meta, Twitter/X, YouTube, TikTok, and the rest of them to the same standard.
Over the years, I have become firmly convinced that social media, at least in its current Big Tech–dominated form, is a gravely toxic institution. Causing mental health problems among youth is not even close to the worst thing these companies have done. There is strong evidence that smartphones and social media are major drivers behind increased loneliness, the worldwide decline in coupling and birth rates, as well as the global surge in racism and fascism. I don’t think we would have seen Brexit or Trump without Facebook—and that’s not the worst thing either. Facebook’s former director of public policy, Sarah Wynn-Williams, recently published a memoir, Careless People, in which, among many other grotesque atrocities, she alleges the company was centrally involved in the Rohingya genocide.
In return, the main positive activity of these companies is destroying the journalism industry by engrossing the advertising money that used to support it.
Now, in many ways a lawsuit is a highly suboptimal way to regulate anything, including social media companies. Rather than a legislature gathering input from the public and drawing up a logical regulatory scheme to be implemented by a well-funded bureaucracy, we have a bunch of states attempting to throw a punch at Meta by sculpting a lawsuit that just might fit through the grab bag of somewhat relevant Supreme Court precedents and laws that have semi-randomly accumulated over time.
But with astronomical penalties being proposed and the possibility of structural remedies—in other words, Meta having to change design features in Facebook and Instagram that are known to addict people—the outcome might end up a fair bit better than a cost of doing business. And with our ever more paralyzed Congress and hypertrophied court system, lawsuits are often the only way to get anything like accountability. There are thousands of other similar lawsuits waiting to see how this one turns out.
So, if this forces Meta to spend a substantial portion of its profits on radically more aggressive moderation, end the infinite scroll and algorithmic feeds, or other changes, good. If it bankrupts the company entirely, even better. If it fails, we can try something else.
UPDATE: Meta and the states settled the case today for $17 billion, but the product changes are significant, including limiting minors to two hours of social media per day and blocks from midnight to 6 a.m.; blocking notifications during school hours and overnight; concealing likes; banning beauty filters in Instagram; and giving an option to deactivate the algorithmic feed that contours content.
.
