Two days after this paper reported that Meta's own accounting of its teen-safety settlement diverged from the states' headline figure — Meta said the deal totaled roughly $18 billion while state attorneys general touted $17.1 billion — the gap has not closed. It has widened, and this time the divergence is happening inside the mainstream press itself. CNBC continues to cite $16.7 billion, the figure tied to the 47 participating states' guaranteed share. [1] WSJ, CNN and AP have converged on $18 billion, Meta's own framing. And on Friday, Fortune ran a piece independently reconstructing the math this paper credited X's AFP wire post with doing first: that whatever the headline number, the settlement functions as roughly a 1 percent tax on Meta's expected revenue over the decade it covers. [2]
Fortune's analysis, written by University of Massachusetts Amherst technology-policy scholar Carolina Rossini for The Conversation and republished Friday, states plainly that "even $17 billion, spread across a decade, amounts to only roughly 1% of Meta's expected revenue over the same period," and that Meta has told investors the deal will not change its financial guidance beyond a single quarterly expense. [2] That is close to identical, in substance if not sourcing, to the framing this paper's Wednesday coverage attributed to X commentary rather than mainstream analysis — evidence that the gap the paper flagged two days ago was not a permanent X/MSM split so much as a lag, with independent scholarly analysis eventually catching mainstream distribution.
What Fortune adds that Wednesday's coverage did not have is a second, sharper frame: this is "a play to box in TikTok and YouTube." [2] Rossini's analysis walks through the settlement's own architecture. Participating states are guaranteed roughly 70 percent of the total, about $12.7 billion, while the remaining 30 percent, about $5.3 billion, is released only if YouTube and TikTok separately adopt matching teen-safety defaults and pay comparable amounts. [2] Meta, in other words, structured part of its own penalty as a recruitment device — telling its two biggest competitors in attention that these rules are coming regardless, and it is cheaper to adopt them together than to face the next round of litigation alone. There is also a defensive logic: if the new defaults make Meta's products less profitable, forcing the same constraints on TikTok and YouTube prevents Meta from absorbing that cost while its rivals do not.
Rossini's piece identifies where the real stakes sit, and it is not the dollar figure at all. The settlement requires teens under 18 to get a default two-hour daily time limit across Instagram and Facebook, an overnight block from midnight to 6 a.m., muted notifications during school hours, hidden like counts, and stronger age-detection systems, among other product changes. [2] But Meta has committed to the strongest of those terms, the daily limit and the overnight block, for only five years of the ten-year agreement; the fuller ten-year version with stricter defaults only kicks in if YouTube and TikTok sign on. [2] An independent auditor reviews Meta's compliance annually, but also only for five of the ten years the deal covers. [2]
That mismatch, five years of guaranteed teeth against a ten-year settlement, is the detail Fortune's piece treats as the actual story, and it is a frame the settlement's own dollar-figure confusion has crowded out of most coverage this week. Rossini poses the open questions directly: who selects the auditor, what information the auditor can access, whether independent researchers can reproduce the findings, and what counts as noncompliance if Meta follows the letter of the agreement while redesigning around its edges. [2] None of those questions turn on whether the settlement is worth $16.7 billion, $17.1 billion or $18 billion. They turn on whether a decade is long enough to find out if two-hour limits and hidden like counts actually change how teenagers use the apps, or whether Meta's competitors will even join the second half of the deal at all.
The financial confusion, in that light, functions less as a reporting failure than as a distraction each side has an incentive to let run. Meta gets to cite the larger $18 billion figure in its own announcements, emphasizing generosity; the states get to cite $17.1 billion or $16.7 billion, emphasizing accountability extracted. Both numbers describe the same document. Neither answers the question Fortune's analysis says actually matters, which is whether product redesign, as a legal remedy, produces measurable behavior change or merely produces another round of settings nobody adjusts.
-- THEO KAPLAN, San Francisco