Eight days into a trial that was supposed to run until October, Meta blinked. On Wednesday, August 26, California Attorney General Rob Bonta and a bipartisan coalition of 51 attorneys general announced a proposed consent judgment that ends the landmark child-safety case in Oakland federal court and commits Meta to pay states up to $17 billion over ten years — while rewriting, by court order, how Instagram and Facebook behave for every user under 18 in the United States. The settlement landed midway through week two, after whistleblower testimony had already gone in and before Mark Zuckerberg was expected to take the stand.
The money is the headline, but the injunction is the story. Under the agreement, teen accounts get a default daily time limit of two hours across Instagram and Facebook combined, which only a parent can turn off. The apps go dark from midnight to 6 a.m. by default. Push notifications are silenced from 10 p.m. to 7 a.m. and again during the school day — 8 a.m. to 3 p.m. on weekdays from August 15 to June 15 — with carve-outs for direct messages and account-security alerts. Full-screen "productive pauses" interrupt continuous scrolling after 15 minutes, then again at 60 and 90. Like counts disappear for minors, and cosmetic-surgery filters are banned outright for that cohort. Teens get an option for a non-personalized feed with no recommendation algorithm behind it, plus a switch to kill autoplay. Meta must answer 90% of teen reports of harmful content within six hours, deploy age-assurance technology to eject users under 13, and submit to an independent auditor with a direct line to the attorneys general.
"Today, we have secured a settlement with Meta that will make social media less dangerous for our kids and make a world of a difference for children and their families," Bonta said in announcing the deal. "We are talking about time limits, stopping notifications during school, a block on the app during critical overnight hours, bans on plastic surgery filters, and so much more." Washington, D.C. Attorney General Brian Schwalb was blunter about the underlying allegations: "Meta intentionally exploited kids for profit and then lied about it, claiming its products were safe when its own internal research confirmed the platforms were addictive and harmful." He added a warning aimed past Menlo Park: "It will not be the last."
Meta admitted no wrongdoing, and its lawyers spent two weeks disputing that its apps cause the harms alleged. But the company's framing is revealing. Chief Legal Officer C.J. Mahoney called the terms "the right path forward for our whole industry, but this framework will only work if all our peers join us." That is not rhetoric — it is the financial architecture. Only about 70% of the payout, roughly $12.7 billion, is guaranteed. The remaining 30%, about $5.3 billion, is released only if TikTok and Alphabet's YouTube adopt comparable one-hour daily limits, night mode and age assurance, and each agrees to pay states roughly $5.3 billion of their own. The injunctive terms escalate on the same trigger: the two-hour cap drops to one hour and the overnight block widens from midnight–6 a.m. to 10 p.m.–7 a.m. only if rivals sign on. In effect, Meta has written its competitors a $5.3 billion invoice and handed it to 51 attorneys general to collect.
Why It Matters
Regulators have spent a decade failing to legislate platform design. Congress never passed a federal kids-online-safety bill; state age-verification laws have been repeatedly enjoined on First Amendment grounds. A consent judgment, entered by Judge Yvonne Gonzalez Rogers of the Northern District of California, does what statutes could not: it sets defaults. The lesson product leaders across the AI and consumer-tech industry will take from Oakland is that engagement optimization is now a provable liability — and that internal research showing harm is the exhibit that ends the trial. Instagram's "Take a Break" feature became a centerpiece precisely because Adam Mosseri conceded on the stand that it saw little uptake until Meta made it a default in 2024, nearly three years after launch. Opt-in safety is no longer a defense.
The comparison state officials keep reaching for is Big Tobacco in the 1990s, and the parallel holds uncomfortably: the money is real, the behavioral change is bounded. The two flagship protections — the two-hour cap and the midnight block — run five years, not ten, under Section II.B.1 of the executed agreement. The auditor's term ends 120 days after its fifth annual report, leaving the back half of the decade unmonitored. The age-assurance standard permits Meta's in-house detection to misclassify up to 14% of 16- and 17-year-olds as adults in year one, tightening to 10% in year two — and Meta has a full year to implement it, six months for general compliance. Against $200.97 billion in 2025 revenue, the guaranteed $12.7 billion spread across ten annual installments is roughly three weeks of sales. Investors read it that way: Meta shares rose about 4.4% in premarket trading Wednesday.
Not everyone is celebrating. Kate Ruane, director of the Free Expression Project at the Center for Democracy & Technology, warned of "significant risks to everyone's privacy and free expression rights online, especially in the ways this settlement will subject all users to invasive age assurance." Age assurance at Meta's scale means everyone proves their age — adults included.
What to Watch
Three things. First, whether Judge Gonzalez Rogers enters the consent judgment as written; nothing takes effect until the first business day after she does. Second, whether TikTok and YouTube take the bait — that $5.3 billion contingency is all-or-nothing, forfeited permanently under Section VI.D.3 if the conditions go unmet for ten years. Third, the litigation Meta did not settle: Florida, New Mexico and Texas sit outside this deal, and more than 3,300 personal-injury cases from families and school districts remain consolidated in federal court in California. Meta bought an exit from one courtroom. The rest of the bill is still open.
“Meta intentionally exploited kids for profit and then lied about it, claiming its products were safe when its own internal research confirmed the platforms were addictive and harmful.”— Brian Schwalb, Attorney General, District of Columbia