Consensus Baseline
The facts every account agrees on.
On Wednesday 21 October, District Judge Francis Mathew in Santa Fe ordered Meta to pay New Mexico $25bn. A jury had found in September that 26 of 29 statements by Meta and its executives about user data, privacy and content moderation were misleading, amounting to more than 43 million violations of the state's Unfair Practices Act. At the maximum of $5,000 per violation, that came to about $219bn; the judge reduced it to $25bn. The state had asked for $35bn to $40bn, and Meta for no more than $3.45bn. The money goes to New Mexico's education system, with interest of 15 per cent a year until it is paid. The case was filed in 2021 and grew partly out of the Cambridge Analytica affair. Meta says it will appeal.
Disputed Claims
The points the accounts contradict each other on.
The numbers are the accounts this piece is built from, listed in full at the foot of the page. Tap one to jump to it.
The case penalises speech.
Meta's chief legal officer calls the suit a tax on speech, and free-speech groups warn it lets states pressure platforms over what they leave up. The judge's findings frame the violations differently: Facebook told users things about its data practices and its own moderation that were false.
New Mexicans were misled or harmed by the statements.
Violations were counted by the number of Facebook users or residents exposed to each statement. Meta argues the jury made no finding that any New Mexican saw or heard the statements, or was harmed by them. The judge cited the size of the audiences reached and the involvement of senior leadership.
Perspectives
How each account frames the same story.
Deception, Plainly
The state and the judge treat this as an old-fashioned consumer fraud case that happens to involve a tech giant.
"Facebook promised users that they controlled their data. This was false," Judge Mathew wrote. He found Facebook's promised investigation after Cambridge Analytica "willfully ineffective", and pointed to the duration of the conduct, the involvement of senior leadership and the "clear intentional and systematic nature" of the misstatements. Those findings, he said, justified the maximum penalty per violation before he reduced the total. Attorney General Raúl Torrez called it a message to Silicon Valley. The state's lawyer, Randi McGinn, was blunter: the court should speak to Meta in the only language it understands, which is money and its share price.
A Tax on Speech
Meta and its defenders say the real target is how the platform handles what users post.
Part of the verdict concerns what Facebook said about hate speech and misinformation, and that is where the free-speech objection comes in. The court found that Facebook made repeated public statements about its ability to find and remove such material while knowing they were deceptive, and that an internal system known as XCheck let posts that broke its rules stay up. C.J. Mahoney, Meta's chief legal officer, says the lawsuit "seeks to impose a tax on speech" and makes doing business in New Mexico all but impossible. Casey Mattox of Stand Together called it a major threat to free speech, and the American Center for Law and Justice said the attorney general should not dictate what Americans can say online. In court, the state argued that Meta dropped third-party fact-checking in January 2025 to please the incoming Trump administration.
The Arithmetic of 43 Million
Looks at the number that drives the whole penalty, and why it will be the centre of the appeal.
Violations were counted per user or resident exposed to each misleading statement, which is how 29 statements became 43 million violations. Meta's lawyer Matt Nicholson told the court that the jury never found any New Mexican actually heard the statements or was harmed by them, and that $40bn would "shock the conscience". The company argued that anything far above $3.45bn would breach constitutional limits on excessive penalties. The $25bn figure is more than twice New Mexico's annual budget, according to the Daily Caller. The judge's own decision to cut $219bn to $25bn shows he thought the maximum was too much. Meta also argues that an appeal could cut the number of violations, which would shrink the base for the entire calculation.
The State That Didn't Settle
Places the ruling in New Mexico's separate strategy of litigating against Meta rather than joining multistate deals.
Nearly every other state settled with Meta over child safety, including about $459m tied to Cambridge Analytica claims. New Mexico stayed out, and that is why this trial happened. It is the state's third win against Meta this year, after $375m in March and $567m in August in a separate child-safety case. Meta has asked to post a $1.8bn bond to pause collection of those earlier sums during its appeal; the state did not oppose that, though it does oppose Meta's request for a new trial. According to a report on the judgment's text, the new order makes no provision for pausing payment or posting a bond, and a ruling on the state's request for an injunction is on hold.
Divergence
Where the accounts part company, and on what.
Each side describes part of the judgment accurately. The state's case is about deception: it says Facebook described its data practices and its moderation in ways it knew to be false. Meta's objection has substance too, because part of the penalty concerns statements about moderating speech, and the state's own argument about the end of fact-checking shows how close the case comes to policing editorial choices. The most exposed part of the judgment is neither of those arguments. It is the count. Because violations were tallied by exposure, with no finding that anyone was misled, the size of the penalty rests on a counting method that Meta's lawyers are already attacking and that the judge himself scaled back by nearly nine-tenths. Two things remain open: whether the 43-million figure survives appeal, and whether other states start pursuing platforms over their own descriptions of moderation.
Sources
Outlet names appear only here, never in the layers.
This article synthesises publicly available reporting for informational purposes. The perspectives summarised are paraphrased from the listed publications and reflect their respective editorial framing. Plural does not independently verify, endorse, or adopt the underlying factual claims, and provides this information on an 'as-is' basis without assuming liability for the accuracy of the original sources or the content of external links.
This summary represents an attempt at neutrality and may still contain bias; it is not presented as absolute truth.
AI is used exclusively to read and compare the independent reports behind each story, and to do nothing else. Programmatic guardrails ensure the system cannot introduce external details, infer motives, or express opinions. Automated testing immediately rejects any draft that violates these rules; failing drafts are discarded, never salvaged.
What survives this process is merely a draft, not a published piece. A human editor then scrutinises the text line by line against the original sources to minimise bias and verify accuracy. No content reaches a reader any other way.