Product4 distinct publishers3 min readPublished
Annualised, the money is well under one percent of Meta's revenue. The product defaults filed in Oakland are the part every consumer app with teen users will be measured against.
The Product Desk · Product desk
Compiled by The Product DeskSomething wrong?How this is made
A default is a spec, and this one is unusually precise for a document drafted by lawyers. Under-18 accounts get a two-hour cumulative daily allowance across Facebook and Instagram, with time on both apps counted together, including on additional accounts Meta detects [4]. Prompts appear after every fifteen minutes of continuous use, and again when daily totals hit 60 and 90 minutes [7]. Run those rules together and a teen who spends the whole allowance can meet as many as eight interruptions before the cap closes [4].
The engineering that is hard to screenshot sits underneath: cross-app accounting and multi-account detection mean linking identities across two properties, and every number in the filing depends on knowing which accounts belong to minors. Meta says it is investing in stronger technology to catch accounts that may belong to under-13s and to identify 13-to-17-year-olds who supplied an adult birthday [20]. It also renewed its call for app stores to verify age and collect parental approval before a download [19]. That is the load-bearing dependency, and Meta would prefer somebody else owned it.
The money is structured to be undramatic. Seventy per cent, about $12.7 billion, is paid in annual instalments over ten years [10], which works out at roughly $1.27 billion a year [1]. Against 2025 revenue above $200 billion [3], that is on the order of 0.6 per cent of a single year's takings [2]. New Jersey's attorney general, Jennifer Davenport, said her state could receive at least $525 million [16].
The remaining 30 per cent is where the interesting drafting is. Meta pays that $5.3 billion only once YouTube and TikTok adopt 15 teen-safety measures designed with the 52 attorneys general, and agree to pay the states another $5.3 billion between them [11]. If they do, Meta's own obligations deepen rather than lapse: the five-year commitment on the cap and Night Mode becomes ten years, the daily limit drops to one hour, and the blackout starts at 10 p.m. [13]. Meta says the protections "will only be truly effective if we work with our peers" [14]. In effect it has bought an option on industry-wide rules and pays the premium only if the option is exercised.
Read the same clause from the other end and the sunset appears. Without peer adoption, the two-hour cap and the midnight-to-6 a.m. blackout are guaranteed for five years [13], and Meta denies wrongdoing throughout [17]. Independent auditors and the participating states are expected to assess both implementation and effectiveness [18], which is the difference between a press release and a covenant.
One more discrepancy worth carrying: the two accounts of the softer settings do not match. PCMag describes hiding like and reaction counts as something teens can choose [8]; Digital Trends lists like counts as hidden by default for younger users [9]. On a settlement whose whole thesis is that defaults are the product, that gap is not cosmetic. The trial began in Oakland last week and the attorneys general filed the deal in the same court on Wednesday morning, with Judge Yvonne Gonzalez Rogers expected to approve it [15].
Ranked by verification strength, evidence, and original report placement.
For users under 18, Meta will limit access to Facebook and Instagram to two hours per day; parents can turn the limit off, and time on both apps counts toward the total, including where Meta detects that someone has multiple accounts.
PCMag reports teens can choose a non-algorithmic feed as their default, turn off autoplay, and hide the number of likes and reactions on posts.
Digital Trends lists hidden likes as a change: like and reaction counts will be hidden by default for younger users.
The remaining 30%, about $5.3 billion, will be paid once YouTube and TikTok adopt 15 teen-safety measures Meta designed with the 52 attorneys general and agree to pay the states another $5.3 billion, split equally between the two companies.
Digital Trends reports Meta stated the 30% tranche of about $5.3 billion will only be paid if YouTube and TikTok implement similar limits and also pay about $2.5 billion each.
A Night Mode will block the apps between midnight and 6 a.m. local time, preventing teens from posting or viewing Feed, Stories, Explore or Reels.
Follow any of these and your For You feed starts watching them — no settings page required.
Evidence-backed comparisons of source perspectives and observed adoption signals. Read the methodology
Which Builder, Operator, and Investor concerns the observed source mix emphasized—not a truth score.
Evidence, demonstrated adoption, hype gap, incentives, and confidence are assessed independently, each on its own current evidence. How these are measured.
Detailed and consistent on product terms, thin and contradictory on numbers, single-origin overall
Four publishers describe the same product defaults with close agreement, which raises confidence in the spec itself. But all four are reporting Meta's own announcement, blog post and open letter rather than the filed settlement document, and they contradict each other on the headline amount (approximately $18 billion versus up to $17.1 billion), the number of participating states (52 AGs across 47 states versus 29 states), the expanded night window (ending 6 a.m. versus 7 a.m.) and the per-company peer payment. Only PCMag reaches outside Meta, citing The New York Times for the court posture. Two of the six items are duplicate republications, so the apparent breadth overstates independent sourcing.
Filed and disclosed, not yet in force or implemented, zero peer uptake
What exists today is a filing and an announcement: the settlement was lodged in Oakland with approval still expected, the defaults apply only pending judicial approval, and no source reports a single feature live for any teen account. The strongest concrete adoption signal is financial and procedural rather than product: Meta expects to book an approximately $10 billion legal expense in Q3 2026, and an annual independent compliance audit is provided for. The conditional tranche, the one-hour cap and the earlier night window all depend on TikTok and YouTube joining, and neither has responded in any cluster source.
Headline dollar figure and industry-standard framing run ahead of the enforceable, unimplemented substance
Coverage and Meta's own language present a new industry standard, but the enforceable core is narrower than the framing: about 0.6% of annual revenue in yearly cash, a five-year floor on the two key limits, parental override on the cap, and the strictest provisions contingent on rivals Meta cannot bind. Digital Trends' caution that effectiveness remains the open question, and PCMag's note that none of it matters if teens misstate their age, are the honest counterweights. The gap is positive but moderate rather than severe, because the product defaults themselves are specific, dated and auditable rather than vague.
Announcement authored by the defendant, with structural incentives to pressure rivals and shift the burden
Every substantive fact in the cluster originates with Meta's announcement while Meta denies wrongdoing and avoids a trial already underway. The settlement structure gives Meta a direct financial incentive to lobby TikTok and YouTube into matching terms, since roughly $5.3 billion is withheld unless they adopt 15 measures Meta designed and pay in themselves, and Meta pairs that with a renewed push to move age verification onto app stores. On the other side, attorneys general have their own incentive to publicise large per-state figures, as with New Jersey's at-least-$525 million. Independent auditors and the states are meant to check effectiveness, which partly offsets the self-interest but has produced nothing yet.
High confidence in the product spec, moderate in the money and low in outcomes
Convergent, granular reporting from four publishers within hours of the filing makes the teen default set reliable at the level of specific hours, thresholds and toggles. Confidence drops for the financial and procedural layer, where the cluster contains three mutually inconsistent settlement descriptions, no primary filing text, duplicate republications inflating apparent corroboration, and no independent confirmation that the court has approved anything. Confidence in effects on teen behaviour or on Meta engagement is effectively nil, since no source measures either.
security
Meta's $17.1B settlement turns teen safety into an audited product spec3 distinct publishers
product
Meta's $18bn settlement is a product spec, and $5.3bn of it is aimed at TikTok and YouTube1 distinct publisher
product
Meta's under-13 data practices go to a jury: 29 AGs, COPPA, and a porous age gate1 distinct publisher
invest
Meta's record child-safety settlement buys off a $200bn trial for about 6 cents on the dollar1 distinct publisher
Distinct publishers with included, body-backed reporting in this cluster.
dexerto.com
2 articles · August 26, 2026
digitaltrends.com
1 article · August 26, 2026
pcmag.com
1 article · August 26, 2026
techcrunch.com
2 articles · August 26, 2026