What Meta’s $17-billion settlement revealed about the harms of social media

After months of hype and mere hours in front of the jury, California’s blockbuster suit against Meta ended on Wednesday with a sweeping $17-billion settlement and 130 pages of promised changes to the platforms.

Witnesses at the blink-and-you-miss-it Oakland trial echoed much of what had already been heard this year by juries in Los Angeles, Santa Fe and Nashville.

But the evidence presented during the proceedings still shed new light on social media’s seedy underbelly.

Internal documents show Meta knew for years that its youngest users were the most likely to be targeted by bullies, extortionists and sexual predators.

Arturo Béjar, an ex-Meta employee, described girls being threatened with rape, boys whose feeds were flooded with snuff films, and middle schoolers inundated with sexually explicit images. Vulnerable children were algorithmically force-fed images of starving bodies and hanging teddy bears, the former engineer said while testifying for plaintiffs in the lawsuit.

Yet the company excluded the majority of those harms from its public reports, designing its metrics so that many of the worst encounters were never captured.

“Whenever Mark Zuckerberg was asked about harm, he would immediately start talking about ‘prevalence,’” Béjar said, using Meta’s term for violations of its company policy. “We found the harm was 100 to 400 times what was being accounted for by prevalence.”

Many of those harms were graphic.

“The body falling and being cut at the last frame, or bones being broken, that’s content people say is violent and distressing but does not fall in [prevalence],” Béjar said. “You experience the full harm of the video even though you don’t see the thing that violates policy.”

That testimony came just after opening statements Aug. 18 . A week later, former Meta data scientist George Volichenko told jurors he and his team were forbidden from making safety settings the default for teens amid fears it would shrink the amount of time that young people spent on the app — the project’s stated goal.

During his trial testimony, Instagram’s director of product design, Francesco Fogu, reviewed an internal chat where he groused that company lawyers had scrubbed key data about teen exposure to suicide, self-injury, and eating disorder content from a presentation to top brass, over worries those executives might later be compelled to testify about them.

“At this point I feel we are purposefully omitting stats that could help them properly assess the situation,” Fogu wrote in the July 2023 exchange, where he noted children were more than twice as likely as adults to be served dangerous content and suffer bullying and harassment.

“These are pretty strong and could change the outcome of the discussion,” he wrote. “Does legal not want us to share them at all?”

In the trial’s final minutes, Instagram boss Adam Mosseri was confronted with a 2021 email in which he excoriated underlings for proposing new child safety features on the heels of his testimony to Congress.

“Announcing this after my testify is worse than before, and we talked this,” Mosseri wrote, according to the email read in court.

It was, experts agreed, a bad look for the company.

“The evidence wasn’t getting better here for Meta,” said Joseph McNally, former acting U.S. attorney for the Central District of California and an expert in “technology-related harm.”

It was also a big wake-up call for families, who may not yet know the extent of the horrors children have faced on the apps.

“Parents, teachers and others are much more educated having had this litigation publicly play out over the past year than they were two years ago,” McNally said. “They’ve then heard and seen through the news coverage that those incidents were corroborated by admissible evidence in court.”

Angst over Meta’s public reputation likely played a role in the decision to settle so early, experts said.

“Who wants to be accused of abusing children?” said Carl Tobias, a professor at the University of Richmond School of Law.

But while the settlement sidesteps any admission of wrongdoing by the company, it does compel Meta to combat problems few parents knew existed, in terms far starker than those it has articulated before.

“You’re notching some big wins with this settlement, which would have been very difficult to achieve at trial and even more difficult to sustain on appeal,” said Peter Jackson, a privacy and cybersecurity attorney in Los Angeles. “Parents are now aware months ahead of time that these controls will exist.”

Under the agreement, Meta must flag parents whenever accounts it has identified as likely to solicit youngsters for nudes and then leverage those images for blackmail start messaging their kids. It will also send alerts if kids interact with posts about suicide, eating disorders and self-harm.

Some parents might wonder why such policies had to be litigated at all. But legal experts said the agreement reflects what was likely weeks, if not months, of careful backroom horse-trading.

“This settlement has been negotiated for a very long time,” Jackson said.

In a sudden about-face, Meta published an open letter to its competitors on Wednesday urging them to adopt sweeping changes the company itself had dismissed as commercially untenable and technologically impracticable during opening statements just days earlier.

“These protections will only be truly effective if we work with our peers — TikTok and YouTube — to put the same measures in place,” the letter said.

Child safety experts lauded the settlement and the accompanying pressure campaign as a giant leap forward in the effort to regulate an industry that until this year appeared all but impenetrable to reform, either from the courts or legislatures.

But not everybody is cheering.

“This settlement agreement is basically the government coercing Meta into censoring users,” said Ari Cohn, free speech counsel at Tech Freedom. “These are things that the government could never legislate. The government could not set a bedtime for the 1st Amendment or limit free speech to two hours a day.”

Although children are more vulnerable than adults, they have the same right to free expression, he said. Efforts to protect them will inevitably spill into censorship of all users — now on the orders of the government.

“Would Meta have actually agreed to this if not for the government wielding its extreme authority demanding gigantic penalties?” Cohn said. “As satisfying as it may feel to some people to see Meta get socked by this, at the end of the day it’s going to come back and bite you.”