Story
August 26, 2026
Meta’s $18 Billion Child-Safety Deal Is Huge—But Critics Say the Hardest Fixes Are Missing
Meta’s proposed nationwide teen safeguards and up-to-$18 billion settlement mark an unprecedented concession. Yet New Mexico and child-safety advocates say the deal leaves key dangers—from predator contact to AI chatbots and algorithms—largely untouched.
Meta has agreed to its biggest-ever child-safety payout and a sweeping reset of teen protections on Facebook and Instagram. But the settlement’s scale has not ended the argument over whether it actually confronts the platform features critics say put children at risk.
The proposed deal, which still requires court approval, follows a federal case brought by a coalition of states alleging Meta designed its apps to be addictive for young users. In opening arguments shortly before the settlement, California Deputy Attorney General Megan O’Neill described Meta’s model in four words: “hook” users, “hold” them, “harvest” their data and “hide” the truth; she added that it was “especially bad for kids.”1
Meta says it will pay up to $18 billion over a decade. About $12.7 billion is guaranteed, while roughly $5.3 billion depends on TikTok and YouTube adopting comparable limits and contributing matching payments. The agreement would impose default two-hour daily limits for under-18s, midnight-to-6 a.m. blocks, school-day notification curbs, hidden like counts, reduced autoplay and expanded age verification.1
For Meta, the bargain is also an industry challenge. Chief Legal Officer C.J. Mahoney said teenagers “move fluidly across dozens of apps,” arguing that the framework needs TikTok and YouTube to follow suit.2
Yet Raúl Torrez, New Mexico’s attorney general, said his state’s earlier court victory went further. He welcomed the nationwide momentum but said officials had hoped for “a direct ban on romantic and sexualized AI chatbot interactions with minors” and tougher protections against adults approaching children in private messages.2
Advocacy group Fairplay struck a similar note: the deal is a “watershed moment,” it said, but does not switch off by default the recommendation systems it says steer children toward predators and dangerous rabbit holes.2 The Center for Democracy and Technology, meanwhile, warned that stronger age assurance could impose privacy and free-expression costs on all users.2
The settlement is historic in dollar terms. Its real test will be whether time limits and parental controls alter the design incentives that drove states to court in the first place.