Time limits and a midnight cutoff: Meta’s new rules for U.S. kids go further, but critics say not enough

Nicklas Akers, California Senior Assistant Attorney General, left, and Deputy Attorney General Megan O'Neill exit court after tech giant Meta reached a settlement to end a landmark trial over teen social media addiction in Oakland, Calif., Wednesday, Aug. 26, 2026. (AP Photo/Noah Berger)

OAKLAND, Calif. — Kids under 18 will soon see a host of changes to their Instagram and Facebook experiences thanks to a historic US$18 billion settlement Meta Platforms reached Wednesday with 48 U.S. states.

Besides the money, which will be paid out over a decade, Meta has agreed to strengthen its age estimates to keep kids under 13 off its platforms and teens in age-appropriate accounts, enact a hard limit of two hours a day for users under 18, block kids from the apps from midnight to 6 a.m., and hide “like” counts, among other changes.

Meta said the new safeguards will only apply in the United States, with most in place for 10 years.

While acknowledging the changes go further than Meta has ever gone when it comes to child safety, critics and children’s advocates cautioned it will not solve its myriad problems overnight.

Arturo Béjar, a former Meta engineering director, said the settlement is a “significant milestone,” but parents should not see Instagram as suddenly safe for children.

“The agreement has a big problem in that it allows Meta to define harm,” Béjar said in an interview Wednesday. “It’s one thing to say, ‘Yeah, you only get like two hours of alcohol or two hours of cigarettes a day,’ but it’s still as bad for you because of what’s getting delivered.”

Béjar, the first witness to testify during the federal trial cut short by the settlement, would like to see Meta address the effect of content recommendations to teens, which he says can prove addictive. While teens might put their phone down when they hit the two-hour limit, they can be distressed about it because the way the platform is designed leaves them craving more.

Meta said people over 18 also can adopt versions of the new controls individually, such as daily time-limit reminders, but there will not be an option for them to opt into the overall settings for what teens experience on the platform.

Stronger age checks

To protect kids, first Meta has to know who they are. While the company has already been using artificial intelligence to determine if kids are lying about their age — either to sign up if they are under 13, which is not allowed, or to use adult accounts when they are 13 to 17 — it did not always work.

Under the proposed settlement, Meta agreed to strengthen its technology to check kids’ ages, using its own as well as third-party tools, with regular outside audits on how well it is working. The agreement also includes specific goals around false positive rates — that is, minors who are identified as over 18. And if a user is identified as under 13 and kicked off the platforms, Meta will check the ages of their friends, too.

“The age assurance section is reasonably good because it has measurement, has independent testing, has goals,” Béjar said. “Because I think Meta should be measured on effectiveness, not effort.”

Meta requires users to be at least 13 years old to create an account, in line with federal child privacy law.

In his testimony last week, Béjar said Meta took a “don’t ask, don’t tell” approach on kids under 13 on its platforms.

“Meta has one of the most sophisticated infrastructures in the world to detect fake accounts,” he said. But despite that, he added, there were “no goals, no metrics” to detect and check kids’ ages who were suspected to be under 13.

Some safety features will be on by default

Critics have faulted Meta over many of its safety features that users have to manually turn on, or opt into themselves.

Of the host of safeguards listed in the settlement, many are on as a default. This includes the two-hour time limit, which can only be turned off with a parent’s permission, as well as the night mode that blocks use between midnight and 6 a.m.

Still, under the agreement other features, such as turning off video autoplay or showing posts in chronological order rather than having algorithms decide what to show kids scrolling through their feeds, will be opt-in.

Josh Golin, executive director of online safety nonprofit Fairplay, said the deal appeared to be too focused on offering tools to parents rather than restricting harmful features altogether.

“We are disappointed that the settlement does not turn off by default recommendation algorithms that connect kids to predators and send young people down dangerous rabbit holes,” Golin said.

Questions around default settings came up at the trial earlier this week when a lawyer for the states suing Meta pressed Adam Mosseri, the head of Instagram, on the low adoption rate among teens for the “take a break” feature, which prompts users to step away if they have been scrolling a while.

Very few teens used the feature when Instagram first introduced it, but since launching separate teen accounts in 2024, Meta has made it the default setting for teenage users. Still, if a teen didn’t want to take a break, they could simply swipe away the notification.

It may take time to see the results

Some researchers have found that many past protections Meta has rolled out are ineffective or don’t work as advertised.

“It is a little premature, unfortunately, to tell if exactly all the changes we want to see are going to happen,” said Yaël Eisenstat, director of policy and impact for the Cybersafety Research Center, which recently audited dozens of safety tools offered by both Meta and other social media companies and said most fell short.

Still, Eisenstat said, Wednesday’s settlement is monumental in showing that alternative approaches are possible.

For years, Meta “decided to go against the advice of some of their own internal employees and continue to ignore the fact that these design features were unsafe in search of profit,” said Eisenstat, who also previously worked at Meta, then Facebook, before leaving the company in 2019. The measures outlined Wednesday at the very least show that there are technically-feasible options that can make the platform safer, she said — something whistleblowers and others have already “been saying very loudly.”

Like others, Eisenstat stressed that there was a lot of work left to be done — and that it will require a combination of technical solutions, market forces and government regulation.

“There’s no singular silver bullet for fixing all of this,” she said. “I’m just glad that at least we are catching up a little bit now.”

The settlement, which also includes the District of Columbia and U.S. territories, must still be approved by a judge.

___

Barbara Ortutay and Wyatte Grantham-philips, The Associated Press

Grantham-Philips reported from Chicago.