When Mark Zuckerberg, the founder and C.E.O. of Meta, the company behind Facebook and Instagram, has tried to defend his business to the public, it’s generally gone poorly. (“I think I’m actually well known to be sort of bad at this,” the billionaire confessed, while giving testimony at a trial in February.) Confronted with leaked internal documents showing that Meta’s products made body-image issues worse for one in three teen girls, and that many of them blamed Facebook and Instagram for increases in over-all anxiety and depression, Zuckerberg dismissed his own company’s findings as user “feedback.” This winter, as a plaintiff’s lawyer in Los Angeles presented him with extensive evidence supporting a link between Meta’s social-media brands and the mental health of its teen-age users, Zuckerberg said, more than a dozen times, “You’re mischaracterizing this.”
Such episodes now look like clumsy moments in a decade of dissembling. Last Wednesday, Meta reached a settlement with the attorneys general of forty-seven states, who had sued it for illegally manipulating the attention of children. The settlement will cost the company, which didn’t acknowledge wrongdoing, up to seventeen billion dollars, a striking figure. Even more striking is a detail reported by the Wall Street Journal: during the second quarter of 2026 alone, Meta spent more than two billion dollars on its legal defense.
The trial, held in Oakland, hinged on an argument that, even if social-media providers cannot be held responsible for the content that their users post—a protection established in Section 230 of the Communications Decency Act—they should be liable for the algorithms that shape what users see. Before the settlement, the plaintiffs had made a strong case. The states presented evidence that senior executives at Meta, intent on maximizing their company’s growth, had repeatedly set aside concerns about what its products were doing to its youngest users. A former Meta engineer testified that he had e-mailed Zuckerberg data showing that the chance of a teen encountering violent or graphic content was between a hundred and four hundred times higher than what the company acknowledged publicly.
Beyond the financial penalty, Meta agreed to make measures that could, if spread through the industry, meaningfully change the experience of social media: Facebook and Instagram will remove the “like” count on posts, do more to verify that their users are at least thirteen years old, limit the service to two hours a day for teen-agers, and restrict service entirely for those users during nighttime hours. There were signs, too, that the public pressure on Meta may yet continue. Florida’s attorney general, James Uthmeier, stayed away from the lawsuit, arguing that it did not go far enough. After the settlement was announced, he released a statement saying, “We’ll see them at trial.”
The settlement was quickly compared to the tobacco lawsuits of the nineties, a moment of public clarity during which a ubiquitous product came to be seen as a scourge. But Meta, a trillion-dollar company increasingly invested in artificial intelligence, is so much vaster than RJR Nabisco, and the current settlement applies to only a single facet of its sprawling operations. There has always been an element of denial in the focus on Instagram’s relationship with minors, when so many of the rest of us, of legal age, also scroll obsessively at weird hours. An unforgiving upcoming film about Meta, “The Social Reckoning,” written and directed by Aaron Sorkin, ends with the January 6th riot at the Capitol, an event that had a lot to do with social media but not so much to do with kids.
It would be easier to view the settlement as a breakthrough in the public’s relationship with technology if we, the public, did not keep making the same mistakes. The problem is speed. New technologies are being disseminated ever more quickly, and the machinery of politics and law delivers accountability at best intermittently, and then only at a painstaking crawl. It took more than a decade of worrying about the way teen-agers used Facebook and Instagram for a legal settlement to enforce a change in Meta’s behavior. Meanwhile, the crypto industry was the largest corporate donor in the 2024 election cycle, and this year’s midterms have been shaped by a voter backlash to A.I. data centers, which have cropped up across the country before the public has had a chance to weigh in on a technology that, as its own developers constantly warn, might be an existential threat.
The scale of private capital available to entrepreneurs has made the tech industry a particularly insular space. One story that the Meta settlement chased off the business pages was that of Leopold Aschenbrenner, a twenty-four-year-old former Columbia valedictorian who’d built a forty-five-billion-dollar hedge fund after making a bullish case for A.I. in a viral online essay. At parties, Aschenbrenner would tell friends that he planned to buy up the rights to faraway galaxies. In late July, when he failed to anticipate a tech downturn, thirty-five billion dollars of his investors’ money vanished.
Zuckerberg has always been a less fanciful kind of founder. Like his products, he expands to fill a social container. In the Obama era, he and his wife launched a Bill Gates-style philanthropy that invested in medical research, education, and public health; these days, he’s into mixed martial arts. (As the journalist Vincent Bevins has emphasized, social media itself followed a similar trajectory, in that what began as a tool to galvanize pro-democracy movements became useful to reactionary forces, too.) When interviewed about Sorkin’s first movie about him, “The Social Network,” Zuckerberg said that its creators had misunderstood him: “They just can’t wrap their head around the idea that someone might build something because they like building things.” But that formulation ignores the crucial matter of what those things do to the people who use them.
The social-media age may finally be waning—Facebook and Instagram chastised, investors looking elsewhere, even Donald Trump’s Presidency, after an extended halftime break, now nearing its fourth quarter. But the relationship between tech and society has not really been altered, so different products continue to follow a familiar pattern: a cloaked and supercilious rollout, a sudden ubiquity, a panicked public response. Two decades on, we are still stuck in versions of Zuck. The week of the Facebook settlement, the cover of Time featured Sam Altman and Greg Brockman, two of the founders of OpenAI, making a defensive case for their product. The headline read, simply, “Trust Us.” ♦
