Advocates for children tell WIRED that Meta’s legal settlement to address concerns about teenagers’ use of social media proposes important and overdue feature changes that invite more parental oversight than ever before. They also, though, say it puts a significant burden on caregivers and unreliable age-assessment systems.
Rather than go through a lengthy federal civil trial—and possible appeals process—in Northern California over the next five weeks, Meta instead on Wednesday agreed to a proposed settlement with a group of state attorneys general, who were suing the social media company for alleged harms against kids on its apps. The settlement includes a payout of up to $16.7 billion overall to states and significant changes to the way Instagram and Facebook work for teen users.
US district judge Yvonne Gonzalez Rogers approved the settlement on Wednesday.
The settlement fee, paid out in yearly installments over 10 years, would be a rounding error for Meta, which raked in more than $200 billion in revenue last year. Meta shares fared well on Wednesday as investors cheered the pending removal of perhaps the biggest of many teen lawsuits targeting the company.
But the proposed changes to Meta’s flagship apps, which include forced time limits, notification blocks, and a chronological feed option for teens, would be some of the most significant changes Meta has made as a result of the ongoing scrutiny of the company’s policies. California attorney general Rob Bonta called it a “watershed moment.”
Still, child safety advocates and mental health experts doubt the effectiveness of Meta’s proposed changes. Digital rights activists, including the Electronic Frontier Foundation, are sounding alarms that the increased age-verification measures that would be built into Meta’s apps for teens would “enshrine Meta's harmful surveillance into law.”
Meta did not immediately respond to a request for comment on concerns raised about the settlement.
What’s Changing
In a big win for the states, parents will now be alerted when teens correspond with suspicious accounts or open a second account, sometimes known as a “finsta,” to furtively use Instagram. In addition, parents will receive periodic updates on their children’s use of Instagram and Facebook, including any attempts to evade parental controls.
Very few of the new measures restrict the types of specific content adolescents can see, with the exception of blocking exposure to extreme makeover filters. But there are several new limits on when young users can scroll their feeds. Access will be blocked overnight and will be limited to two hours during the day, unless parents allow more. From August 15 to June 15 each year, during the US school year, notifications will be muted by default from 8 am to 3 pm.
And users may also opt into a non-algorithmic feed and turn off autoplay of videos, though how Meta plans to make those potentially helpful options palatable to teens was unaddressed. Parents, however, can make those settings the default on their children’s accounts.
“These are common-sense regulations that will do a lot of good,” says Jean Twenge, who wrote the book 10 Rules for Raising Kids in a High-Tech World and testified as an expert psychologist for the states during the trial last week. “Teens should not be using social media in the middle of the night, at school, or for so many hours a day that it crowds out time for talking with friends and family, going outside, exercising, and sleeping.”
One of the changes—hiding “like” counts by default—was first tested among a broad set of users about seven years ago. Despite internal research showing that publicly visible “like” counts could contribute to people engaging in harmful social comparison, Meta only rolled out the ability to hide them as an option. “People didn't really go in and adjust those settings in any significant quantities,” Vaishnavi Jayakumar, a former head of youth policy at Meta, said in a deposition last year.
Carol Vidal, an associate professor of psychiatry and behavioral sciences at the Johns Hopkins University School of Medicine, says the proposed changes are broadly positive. She points specifically to muted notifications at night, for sleep health, and Meta’s proposed limits on plastic surgery filters, noting that “a lot of girls, particularly, are struggling with self-image and comparing themselves online to other people on social media.”
But, Vidal says, measuring the effectiveness of the changes could take years to research.
Why Experts Are Still Concerned
The widespread, immediate responses to Meta’s proposed changes underscore just how complicated it is to regulate social media.
Tanzeem Choudhury, a professor of integrated health and technology at Cornell Tech, the New York City–based research arm of Cornell University, says the proposed changes mostly inject “friction of use.” She points out that measures such as daily limits, nighttime blocks, or muted notifications can be bypassed through secondary accounts or false ages.
So a lot is riding on Meta’s ability to successfully age-gate kids into an experience it calls Teen Accounts. “None of the protections will work if kids and teens can just choose a birth year that makes them 18 or otherwise evade age assurance,” says Twenge, the expert witness.
As part of the deal, Meta agreed to strengthen age assurance but provided few details. “While we'll do all we can to accurately understand someone's age, through both AI prediction technology and steps to confirm someone's age (like ID collection or Yoti's facial scanning), we need app stores to provide us with verified age information from devices,” Meta spokesperson Stephanie Otway says.
Even if successful, age-verification requirements are unwelcome to privacy and security activists. Kate Ruane, the director for the Free Expression Project at the Center for Democracy and Technology, says Meta’s new age assurance tools could pose “significant risks to everyone’s privacy and free-expression rights online.” Electronic Frontier Foundation senior counsel David Greene says that the measures will mean teens have “a lesser ability to exercise their rights to speak, access information and art and culture, associate and form communities, and play.”
Meta also agreed to verify parents more strictly, including limiting the number of accounts each can supervise to hedge against someone running a fake-parent mill. Adam Mosseri, who oversees Instagram for Meta, said on the witness stand this week that at one point less than 1 percent of teens had linked a parent to their account.
It's unclear how Meta will go about significantly boosting that number to get better adoption of its new parental oversight mechanisms. Its proposed system also does not appear to account for parents who are unsupportive of their children, including those who may be gay or lesbian and using social media to find online community. “It could widen existing disparities between children with parents who are actively involved versus those who are less involved or less aware,” Choudhury said.
Meta’s Otway says “the vast majority of the changes in the settlement are on by default, and parents don’t have to do anything” unless they “choose to modify these defaults.” He also notes that teens can set the supervisor of their account to be someone other than their parent, as long as that person is at least 18.
