Meta agreed Wednesday to pay up to $18 billion and impose a two-hour daily limit on teenage use of Facebook and Instagram, resolving claims brought by nearly every U.S. state that the platforms were designed in ways that harmed children. The Aug. 26 agreement also creates default overnight blocks, mutes most school-hour notifications and requires stronger age checks, subject to court approval.

The settlement ends a landmark federal trial in Oakland and a wider group of state cases without an admission of wrongdoing. A court filing says the parties will ask judges to enter consent judgments, making the restrictions enforceable rather than voluntary product promises. Most of the behavioral terms are expected to remain in place for 10 years.

The deal is consequential not because it proves that social media caused a national mental-health crisis—the strongest scientific reviews say that causal evidence remains incomplete—but because it changes the default architecture encountered by millions of minors. It shifts several decisions about time, sleep and content away from teenagers and toward parents, while creating an independent auditing structure intended to test whether Meta implements what it promised.

Two hours becomes the default ceiling

Under the agreement, users identified as under 18 will be limited by default to two cumulative hours a day across Facebook and Instagram. The clock excludes some messaging, settings and long-form video activity, an important qualification to the headline limit. Once the applicable time is exhausted, a teenager cannot restore ordinary access without approval from a supervising parent.

The settlement also blocks most teen access from midnight to 6 a.m. and silences most push notifications from 8 a.m. to 3 p.m. Direct messages and account-security or safety alerts are exceptions during school hours. Meta’s own summary says teenagers will receive prompts after every 15 minutes of continuous use and when daily use reaches 60 and 90 minutes.

Those measures convert familiar parental-control options into defaults that minors generally cannot weaken alone. That distinction matters because optional safety tools often depend on families finding, understanding and activating them. The agreement instead places the burden on Meta to identify teen accounts, apply restrictions and account for teenagers who maintain multiple profiles.

The payment structure rewards wider industry adoption

The financial settlement combines a large guaranteed payment stream with additional money linked to whether competing platforms adopt comparable protections. Reuters reported that Meta will guarantee $12.7 billion over a decade, while roughly $5 billion more depends on similar changes by Snapchat, TikTok and YouTube. A separate Texas agreement pushes the potential total to about $18 billion.

That design gives Meta an unusual financial interest in making the restrictions an industry norm. Meta argues that teenagers move among apps and that limits on one service may simply redirect attention to another. In an open letter, the company urged TikTok and YouTube to accept the same framework, presenting uniformity as necessary for meaningful protection.

The participating states can use substantial portions of the money for remediation, enforcement and youth-focused programs, though allocations and permitted uses vary. California’s share, for example, includes funding directed toward preventing or addressing mental-health and other harms associated with youth social-media use. The payments compensate governments and support programs; they do not establish damages for every individual teenager or settle all private lawsuits.

Age verification becomes the operational test

Every major protection depends on Meta correctly identifying who is a minor. The settlement requires the company to improve systems for detecting children under 13 and to evaluate accounts through an age model, while an independent auditor reviews implementation and enforcement targets. A nominal two-hour ceiling has little effect if young users can enter a false birth date, open another account or move unnoticed between profiles.

Age assurance creates its own privacy tradeoffs. More accurate systems may rely on identity documents, facial-age estimation, account behavior or information associated with connected users. Each method can produce false matches or collect additional sensitive information, so the practical standard is not merely whether a system catches more children but whether it minimizes data collection, explains decisions and offers a credible correction process.

The agreement attempts to address evasion by counting detected multiple accounts toward one cumulative limit. It also restricts certain augmented-reality effects and gives parents additional controls. But enforcement will depend on technical performance that cannot be inferred from the settlement text alone. Public auditor reports and state enforcement actions will provide the clearest evidence of whether the rules reach the users they are meant to protect.

The health evidence is serious but not simple

The legal claims unfolded against widespread concern about youth mental health and nearly universal platform use. A U.S. surgeon general advisory says as many as 95 percent of people ages 13 to 17 use a social-media platform, one-third report using social media almost constantly and adolescents spending more than three hours a day face twice the risk of symptoms including depression and anxiety. Those figures show association and exposure, not proof that platform time caused each reported condition.

A National Academies review reached a deliberately cautious conclusion: social media can produce benefits and harms, but the evidence is insufficient to say it causes population-level changes in adolescent health. Many studies cannot determine directionality. Young people experiencing depression or isolation may use platforms differently, while platform experiences may also influence sleep, self-comparison, harassment exposure and emotional well-being.

That uncertainty does not make design irrelevant. The same review identified persuasive features, endless feeds, reward sensitivity and displacement of sleep or exercise as plausible mechanisms deserving targeted research and safeguards. The American Psychological Association similarly recommends that social-media use not interfere with sleep or physical activity and that adolescents receive developmentally appropriate oversight. The settlement effectively treats those mechanisms as risks worth reducing before every causal pathway is settled.

What the agreement leaves unchanged

The settlement does not eliminate personalized recommendations or targeted advertising, two central components of Meta’s business model. It narrows how teen accounts operate but leaves adults’ services largely intact. It also does not resolve continuing litigation by Florida and New Mexico, according to AP reporting, or automatically dispose of claims brought by individuals, school districts and other nonparticipating parties.

Meta denies wrongdoing and emphasizes that it introduced Teen Accounts before the settlement. Those earlier protections included stricter messaging and content settings, as well as parental supervision tools. The new agreement goes further by attaching judicial enforcement, multistate oversight and financial consequences to specific defaults, but it does not establish that every alleged injury was caused by Facebook or Instagram.

The final terms also require judicial approval. Implementation deadlines generally run from the effective date, so the promised changes will not all appear immediately. The most useful near-term evidence will be whether courts approve the consent judgments, how Meta verifies age without creating new privacy problems, and what the auditor reports about compliance, errors and circumvention.

A national standard built through state litigation

Congress has debated children’s online-safety rules for years without establishing one comprehensive national design code for social platforms. State attorneys general instead used consumer-protection and children’s privacy claims to negotiate operational rules with nationwide reach. New York’s announcement describes the agreement as covering age verification, time limits, overnight access, school-hour notifications and algorithmic-feed choices.

That route has limits: a settlement binds the parties and participating jurisdictions, not the entire technology sector. Yet the contingent payment structure is explicitly designed to pull other major platforms toward the same baseline. If competitors join, the agreement could function like an industry standard; if they do not, teenagers may shift time among services and weaken the practical effect of Meta-only limits.

What changed Wednesday is measurable: Meta accepted court-enforceable restrictions on teen accounts and a payment obligation that could reach $18 billion. What remains unproven is whether those defaults will reduce harmful use, improve sleep or produce better mental-health outcomes. The settlement’s lasting importance will depend less on its headline value than on verified implementation and evidence that the new design changes how young people actually use social media.