Eight days into a federal trial in Oakland, Meta stopped fighting. On August 26, 2026, the company and a bipartisan coalition of state attorneys general filed a proposed consent judgment and a joint motion to enter it, resolving claims that Instagram and Facebook were built to addict children and that Meta lied about it. The money is the headline. The product changes are the story.
New Jersey co-led the case alongside California, Colorado and Kentucky. Attorney General Jennifer Davenport announced that the state will collect at least $525 million and possibly more than $752 million. “As a parent, protecting your kids is always your North Star,” she said. Officials are calling it the largest consumer protection settlement in American history outside the 1990s tobacco settlements, and the comparison is fair.
The settlement ended a trial Meta had already begun to lose
The states sued in 2023, alleging that Meta engineered its platforms to maximize adolescent engagement, understood from internal research what that engagement was doing, concealed those findings and violated the federal Children’s Online Privacy Protection Act by collecting data on users under thirteen. Judge Yvonne Gonzalez Rogers of the Northern District of California had already granted the states summary judgment on the COPPA question in June.
Trial opened on August 18 as a bench trial with an advisory jury, meaning the judge would write the remedy. Instagram head Adam Mosseri testified on August 25, with Zuckerberg scheduled to follow. The states put realistic exposure near $200 billion. Meta argued the penalty theory implied a ceiling around $1.4 trillion, roughly its entire market capitalization. Rather than let a judge pick a number and design an injunction, Meta negotiated both. It admits no wrongdoing, both sides waived appeal rights, and nothing binds anyone until Judge Gonzalez Rogers approves the deal.
What Meta is actually required to do
The obligations fall into six categories.
Usage defaults for minors. Users under 18 face a default two-hour daily limit that only a parent can switch off. It is cumulative rather than per app, so scrolling on both counts against the same two hours, and Meta has committed to applying it across multiple accounts it detects belong to the same person. Night mode blocks access by default from midnight to 6 a.m. School mode mutes push notifications by default from 8 a.m. to 3 p.m., excepting direct messages and security alerts. Continuous scrolling triggers a prompt every fifteen minutes. Direct messaging sits outside all three restrictions.
Design changes for minors. Young users get a feed option that is not ranked by Meta’s recommendation system and the ability to turn autoplay off. Like counts and reaction counts are hidden by default, on their own posts and everyone else’s. Cosmetic surgery filters and extreme makeup filters are unavailable to them entirely.
Age assurance. Meta must deploy enhanced age assurance to keep users under thirteen off the platforms and to identify which accounts belong to minors.
A truthfulness injunction. Meta is barred from making false, misleading or deceptive statements about its platform safety and how its safety features work. This is the sleeper provision, converting future marketing claims into contempt exposure rather than a new lawsuit.
Independent audit. An outside auditor tests compliance and reports to the states annually for five years.
An independent research foundation. Meta must help fund a research body and share consented user data with it so outside scientists can study adolescent well-being without relying on Meta’s own published findings. Since the states’ theory rested on internal research Meta buried, this may be the settlement’s most consequential term.
Most terms stay in place for ten years. The time limit and night mode commitments run five, and stretch to ten only if competitors sign on.
Thirty percent of the money is a lever pointed at TikTok and YouTube
Meta pays roughly seventy percent of the total, about $12.1 billion, unconditionally. The remaining $5 billion comes due only if TikTok and YouTube adopt comparable measures and each pays the states a comparable sum. The conduct terms escalate on the same trigger. If both competitors join, the daily limit tightens to one hour per app, night mode expands to 10 p.m. through 7 a.m., and the framework extends to a decade. Meta published an open letter to both companies the same day, arguing the industry needs an industry-wide solution.
That argument is correct and also self-serving. Teenagers who lose two hours on Instagram do not put the phone down. They move. A limit binding one company and not its competitors redistributes attention rather than reducing it. Meta has now given every state attorney general a funded reason to pursue the other two platforms, and arranged matters so its own remedy weakens if they fail.
Age assurance is the provision that reaches adults
Every protection above depends on Meta knowing which accounts belong to minors, and no platform can verify that some users are under eighteen without evaluating everyone.
That is the part the average adult should read twice. In practice it means more identity signals collected, more inference drawn from behavior and device data and, in some designs, government identification or facial age estimation. The Center for Democracy and Technology warned that a settlement built this way carries real privacy and expression risk for all users. A remedy aimed at adolescent mental health is being purchased partly with adult anonymity, and no consumer was asked about that trade.
No consumer receives a check
The money goes to the states, not to families, paid annually over ten years and allocated by population. There is no claims website and no distribution to individual users.
What each state does with its share is a legislative question. California earmarked its funds for youth mental health. New Jersey has not said how it will spend at least half a billion dollars. If the tobacco settlements are the template, the lesson is uncomfortable, because much of that money drifted into general funds and never reached the harm it was collected to remedy.
People with actual injuries litigate elsewhere, in thousands of personal injury and school district claims still pending in the same multidistrict litigation. Three states stayed out. New Mexico won its own verdict and a separate $567 million abatement order, Texas settled separately for $1 billion, and Florida remains unresolved.
What this changes for you before the court approves anything
Parents should look at the accounts their children actually use rather than wait for defaults to appear. Anything on TikTok, YouTube, Snapchat, Discord or a plain browser is untouched, and direct messaging is carved out even on Meta’s apps. The parental permission mechanism matters too, because a control a teenager can talk a tired adult into disabling at 11 p.m. is not much of a control.
Adult users should expect age checks on products they have used for fifteen years without proving anything about themselves.
Anyone building consumer software should read the consent judgment rather than the coverage of it. It is the most detailed public statement yet of which design choices regulators consider defective as opposed to merely aggressive. Default settings, notification timing, autoplay, algorithmic ranking and the honesty of your safety marketing are all treated as product features carrying legal consequence. That framework will not stay confined to social media, and it will not stay confined to children.
You may also enjoy:
- Should Kids Use AI?
- New Jersey Takes the Lead on E-Bike Safety
- New Jersey Bans the Sale of Sensitive Data and Puts Data Brokers on a Public Registry
- Trump Accounts Are Live. What Every Parent Needs to Know
- Big Tech Loses in Court: What the Social Media Addiction Verdicts Mean
and, if you like what you’ve read, please subscribe below or in the right hand column.