Meta's $18B settlement omits the AI chatbot ban New Mexico already won

Meta's $18 billion settlement with 51 state attorneys general is historic in size, but New Mexico Attorney General Raúl Torrez says it leaves out the strongest protections his state already won in court. Specifically, the nationwide deal does not include a direct ban on romantic or sexualized AI chatbot interactions with minors, nor the tougher safeguards against adults targeting kids in private messages that a Santa Fe jury and judge imposed on Meta. Torrez, whose office secured the first jury verdict against Meta over child safety anywhere in the country, told Fortune he had hoped the national settlement would "echo the full strength" of New Mexico's court-ordered protections. He still called the deal "real progress" and "historic," but his critique is a notable crack in what has otherwise been a unified front of attorneys general.

New Mexico's own tally against Meta now stands at roughly $942 million, built from a March jury verdict of $375 million in civil penalties for 75,000 violations of the state's consumer protection law, plus a judge's $567 million ruling in August that Meta created a "public nuisance" similar to air pollution. The settlement Meta struck this week, still subject to court approval, requires a default two-hour daily time limit, a nighttime block between midnight and 6 a.m., muted notifications during the school day, hidden like counts, a ban on cosmetic-surgery and extreme makeup filters, stronger age verification, and an independent auditor to check compliance for five years. California Attorney General Rob Bonta's office, which led the case, has not yet responded to Fortune's request for comment beyond its public statements.

Child advocacy group Fairplay called the deal "a watershed moment for the growing movement to protect children from addictive and dangerously designed social media," pointing to sleep protections like the nighttime block as "the most significant injunctive relief yet from Meta." But its statement quickly turned to the same kind of gap Torrez flagged. "We are disappointed that the settlement does not turn off by default recommendation algorithms that connect kids to predators and send young people down dangerous rabbit holes," the group said. "In general, the settlement is too focused on offering parents tools rather than restricting harmful features. It also relies a lot on nudging users toward breaks, and we are skeptical that that will be effective. And even the financial penalties, while the biggest Meta has ever faced, are not large enough to fundamentally change Meta's relentless targeting of youth." Fairplay tied its critique directly to a specific ask: a floor vote on the Kids Online Safety Act (KOSA), which has stalled in Congress for years despite support from more than three-quarters of the U.S. Senate.

The Center for Democracy and Technology warned that fixing one problem creates another. "Meta has agreed to implement several changes across its platforms as part of its settlement with 52 state attorneys general," said Kate Ruane, the group's director of the Free Expression Project. "As part of that agreement, Meta is providing tools to help families make their own decisions about kids' online experience and screen time, giving kids and their parents more choices and control is beneficial. But we also see the potential for significant risks to everyone's privacy and free expression rights online, especially in the ways this settlement will subject all users to invasive age assurance and limit all kids' access to content and services regardless of their individual needs." That tension, wanting stronger age checks while worrying about what those checks require of everyone else, has defined the broader fight over kids and social media this year, as platforms turn to facial scans, ID uploads, and other biometric tools to figure out who is a minor. Most Americans don't trust that any of it will actually work, and reporting shows kids find ways around the checks that do exist, including drawing on facial hair to fool age-estimation software.

Phillip Yannella, co-chair of the privacy, security, and data protection practice at Blank Rome, told Fortune the settlement's significance may be less about what Meta agreed to than about what it signals for Washington. "Congress, which hasn't done a thing on privacy in forever, the one issue that they do care about is children's safety, and it does appear like KOSA is moving forward." He was more cautious than the advocacy groups and Torrez about calling Meta's concessions inadequate. "Children's safety advocates and plaintiffs' lawyers are going to take a maximalist view of children's safety, and I would imagine they would suggest there's much, much more that could be done," he said. "But this is a settlement, and sometimes you don't want the perfect to be the enemy of the good. These are steps in the right direction if you're looking at it from a children's safety perspective that weren't there yesterday." Still, he said the settlement could mark a turning point beyond Meta alone: "You could look at this and say this is really the first domino to fall, and there's going to be a lot more changes in this environment, not just for Meta, but for all of them: TikTok and everyone else."

The same trade-off is playing out abroad. Australia, the U.K., and France have all moved toward under-16 social media bans this year, and each has run into the same problem Ruane is describing: verifying a child's age tends to mean verifying everyone's. Meta's Chief Legal Officer C.J. Mahoney, in a statement to Fortune, called on TikTok and YouTube to adopt the same restrictions, saying "teens move fluidly across dozens of apps, we need an industry-wide solution." Neither TikTok nor YouTube responded to requests for comment. Julie Scelfo, founder of Mothers Against Media Addiction, told Fortune the settlement was "great news to wake up to" but stopped well short of calling it enough. For executives watching this space, the lesson is clear: negotiated settlements can move fast, but they may not match the force of a court ruling, and the gap between the two is where the next regulatory battle will be fought.