Meta’s proposed settlement with 47 U.S. states, the District of Columbia and territories would move several teen-safety features on Facebook and Instagram from optional tools to default product restrictions. The deal is valued at up to $17.1 billion over 10 years, although roughly 30% of that headline amount is conditional on other major platforms accepting comparable obligations.
5
17
What would change for teen accounts
For users under 18, the agreement calls for default protections that parents—not teens acting alone—can override in key cases.
A two-hour limit across Facebook and Instagram
Teen accounts would receive a two-hour daily limit combined across Facebook and Instagram. The limit is not a strict ban on all activity: reported terms exclude messaging, settings and qualifying long-form video from the calculation. Parents can turn off the default limit.
17
24
The settlement also includes use prompts, including an alert after 15 minutes of continuous use and additional pauses at longer intervals.
17
Overnight access restrictions
Facebook and Instagram would block teen access from midnight to 6 a.m. by default, unless a parent overrides the restriction. Reports indicate messaging is excluded from the overnight block.
4
9
Quieter school hours and nights
Meta would disable most push notifications for teens during the default school-day window of 8 a.m. to 3 p.m., as well as overnight. The purpose is to reduce interruptions and social pressure during school and sleep hours.
1
5
Hidden popularity metrics and fewer appearance pressures
The company would stop showing teens the number of likes and reactions their posts receive. It would also restrict cosmetic-procedure image filters for minors.
1
32
Stronger age checks—but not necessarily ID verification
Meta must improve age-assurance measures intended to keep children out of age-restricted experiences. However, the reported terms do not require universal photo-ID checks or a single specified verification method, leaving the implementation details significant.
5
30
More parental control and a non-personalized feed option
The settlement would expand parent controls and require an option for teens to use a non-personalized, chronological-style feed rather than a feed ranked around individual engagement signals.
6
25
That is a material choice for users, but it is not a ban on algorithmic recommendations. Meta can still operate personalized feeds for teens who use them.
Why the $17.1 billion figure needs context
The maximum settlement value is up to $17.1 billion, but the guaranteed portion is about 70%—roughly $12.7 billion—paid over a decade.
5
17
The remaining roughly $5 billion is contingent on Snap, TikTok and YouTube accepting substantively comparable child-safety obligations and related financial resolutions. Reuters reported that the contingent phase would involve stricter measures, including a one-hour daily limit per app and wider overnight restrictions from 10 p.m. to 7 a.m.
5
31
In practical terms, the agreement is designed not only to change Meta’s products but also to encourage an industry-wide baseline. Whether that escalation happens depends on competitors and further settlements, not Meta alone.
The main criticism: limits may leave the engagement model intact
Critics have focused on the distinction between restricting specific high-engagement surfaces and changing the broader incentives of social platforms.
The reported exclusion of messaging and video lasting at least 22 minutes from the daily limit has raised concerns that the rules could push usage toward long-form content rather than reduce total time spent on Meta services.
24
35
Likewise, an optional non-personalized feed gives teens an alternative to recommendations but does not remove personalized ranking as the default model. A former Meta whistleblower told Reuters that the settlement did not go far enough to protect teen mental health, even with measures such as hidden reaction counts.
32
The counterpoint is that the deal still requires consequential default design changes: a parent-controlled time cap, a nightly access block, muted notifications, stronger age checks and reduced public popularity metrics.
1
5
17
What the settlement means for Europe
The U.S. settlement does not determine Meta’s obligations in the European Union. But it has sharpened attention on the European Commission’s Digital Services Act investigations into Facebook and Instagram.
The Commission has said it reached preliminary findings that Meta’s platforms were in breach of the DSA in relation to addictive design, and it has separately raised concerns about protections against under-13 access. Meta disputes the allegations. A confirmed DSA infringement can carry fines of up to 6% of a company’s worldwide annual turnover.
40
41
European lawmakers have called for faster enforcement and product changes in Europe comparable to—or stronger than—the U.S. commitments. The Commission has said it remains in dialogue with Meta about protections for children in the EU.
40
46
51
The key difference is regulatory approach:
- United States: a negotiated settlement combines payments with prescriptive defaults for teen accounts, while some money and tougher next-stage restrictions depend on competitors joining.
5
17
- European Union: DSA proceedings focus on whether Meta has adequately assessed and mitigated systemic risks, including risks associated with addictive design and child access. The available record describes preliminary findings and ongoing discussions, not a final EU penalty or final remedy.
40
41
Bottom line
For U.S. teens, the proposed settlement would mean more than wellness reminders: Facebook and Instagram would be required to make time limits, overnight restrictions, muted notifications and parental oversight the default. Yet the agreement does not eliminate algorithmic recommendation, its largest financial component is conditional, and the effectiveness of stronger age checks will depend on how Meta implements them.
5
30
In Europe, the deal is becoming a benchmark for a broader question: whether DSA enforcement will require Meta to change the underlying design of recommendation-driven services, rather than simply add guardrails around them.
40
46