Meta Agrees to Pay Up to $16.7 Billion and Change Instagram and Facebook for Teens

Meta agreed to pay up to $16.7 billion to settle a lawsuit brought by 51 US states and territories that accused the company of failing to protect young users. The settlement was announced, according to WIRED, before even a quarter of a federal trial expected to last 19 days had been completed, shortly before Instagram's chief, Adam Mosseri, was set to testify again.

More relevant than the check, however, are the product changes the company will have to implement, and the precedent the case sets for platform accountability at a global scale.

What came before

Meta had been resisting settlements where competitors gave in. YouTube, Snapchat, and TikTok resolved similar lawsuits outside of court; Meta repeatedly insisted on going to trial. This year alone, it lost state lawsuits in California and New Mexico, racking up nearly $1 billion in penalties. The federal trial that has now ended in a settlement threatened to expose the company to its own estimate of more than $1 trillion in damages.

The 29 states that went to trial alleged that Meta violated federal children's privacy laws by collecting data from users under 13. Four of them (California, Colorado, Kentucky, and New Jersey) argued that the company misled parents with false claims about protections and designed features in an abusive way. Meta denied all the allegations.

During the presentation of the case, former and current employees described a culture of maintaining or growing engagement above all else. An academic psychologist testified that teenagers, especially girls, who spent more hours on social media were more likely to experience depression, contributing to rising loneliness and self-harm over the past 15 years.

The technical detail that matters to product builders

Part of the testimony focused on features that Meta itself developed to curb excessive use, and on their adoption numbers. The Take a Break feature, which reminds teens to limit prolonged Instagram use, had been activated by less than 1% of weekly teen users months after launch. Two years later, adoption was still only 1.8%. Another feature, Quiet Mode, performed almost identically.

This is the point product devs should take note of: a wellbeing feature delivered as a hidden opt-in is practically never used. The settlement corrects this by flipping the logic to mandatory defaults, with opting out conditioned on parental approval.

What changes on Instagram and Facebook

Subject to court approval, the changes include:

• A two-hour daily limit by default for teens, on both platforms; turning it off requires parental approval.
• An exception for long-form content (audio or video of 22 minutes or more) and direct messages, which fall outside the limit.
• Apps blocked between midnight and 6 a.m. by default for teens.
• Notifications muted between 8 a.m. and 3 p.m., from August 15 to June 15, to reduce distraction during school hours.
• Warnings every 15 minutes of continuous screen time, with additional alerts at 60 and 90 minutes.
• Parents can set a chronological feed instead of an algorithmic one and turn off autoplay by default.
• Likes and reactions hidden by default for teens.
• Deactivation of cosmetic surgery and extreme makeup filters for teens.

The whole architecture revolves around safe default + parental control, the pattern the industry has been calling safety by default.

The $16.7 billion asterisk

The amount has an interesting condition. About $12.7 billion is guaranteed and will be paid over 10 years in annual installments. The rest depends on other "core industry members," defined as Snap, TikTok, and YouTube, adopting similar safeguards: a one-hour daily limit, a comparable night mode, and age verification (age assurance) measures. If that happens, Meta's own daily limit drops from two hours to one, and TikTok and YouTube would have to make a combined payment comparable to Meta's remaining balance.

In practice, Meta turned the settlement into competitive pressure: in its own communication, the company called on YouTube and TikTok to adopt the same changes to "set a new industry standard."

What Brazilian developers can take from this

There's no fine here for Brazilian platforms, but the trend this case consolidates matters for those building software in Brazil for a few concrete reasons:

Age verification becomes an infrastructure requirement. If age assurance becomes a market standard imposed by the biggest players, teams that integrate social login, display content, or run recommendations for young audiences will have to think about age estimation/confirmation flows, with all the sensitive-data and privacy costs that entails. In Brazil, this connects directly to the LGPD (Brazil's data protection law) and the ECA (Brazil's Child and Adolescent Statute), as well as the Digital Statute for Children and Adolescents recently signed into law in the country.

Default matters more than feature. The 1.8% adoption number for Take a Break is empirical evidence that a "wellbeing button hidden in settings" doesn't solve anything from a compliance standpoint. Regulators are looking at the product's default state, not at whether the feature exists. For those designing onboarding and settings, the takeaway is that the default itself can become subject to enforcement.

External audit as part of the design. The settlement provides for an independent auditor who can report concerns to prosecutors, in addition to more robust parental supervision tools. Verifiable external accountability tends to require logs, audit trails, and product behavior instrumentation designed from the architecture stage, not improvised afterward.

What remains open

The settlement still depends on court approval by Judge Yvonne Gonzalez Rogers. The contingent $4 billion installment may never be paid if competitors don't sign on. And there's the semantic debate raised by California Attorney General Rob Bonta, who described some measures as a "ban" for teens rather than a mere default adjustment, when in practice parents seem to be able to approve changes from the default anyway. How these safeguards will be implemented technically, and how hard they'll actually be to bypass, is the next chapter to watch.