A Santa Fe judge ordered Meta Platforms to deposit $567,000,000 into a five-year abatement fund on August 6, 2026, on top of the $375 million civil penalty a jury imposed in March, while refusing the State of New Mexico's request to redesign the recommendation algorithms that carry advertising on Facebook and Instagram.

Chief District Court Judge Bryan Biedscheid of the First Judicial District Court entered findings of fact, conclusions of law and a judgment in State ex rel. Torrez v. Meta Platforms, Inc., case number D-101-CV-2023-02838, on the afternoon of Thursday, August 6, 2026. The decree closes a two-phase proceeding that began with a jury trial running from February 2 to March 24, 2026, and continued with a bench trial held between May 4 and May 22, 2026.

The Phase 1 jury concluded that Meta committed 75,000 violations of the New Mexico Unfair Practices Act and assigned each violation the statutory maximum of $5,000, producing a verdict of $375,000,000 recorded in an order correcting the verdict form on March 26, 2026. Phase 2 addressed a separate question: whether Instagram and Facebook constitute a public nuisance in New Mexico, and what abatement would follow.

Biedscheid concluded that they do.

Where the $567 million goes

The court ordered the money split across five categories. Treatment absorbs $420,000,000, the overwhelming majority. Screening and Assessment receives $90,000,000. Awareness and Prevention takes $33,000,000, Referral, Linkage and Coordination $15,000,000, and Implementation, Continuous Quality Improvement and Evaluation $9,000,000.

The figure derives from an abatement plan prepared by Dr. Dana Weiner, an expert in child welfare policy, and costed by economist Dr. Kelly O'Donnelly. Biedscheid did not adopt their numbers intact. Weiner had proposed a 15-year abatement period, arguing that duration was "necessary to effectively implement a sustainable infrastructure and workforce to address and begin to mitigate social media-related harms." The court cut that to five years, reasoning that a 15-year plan "would necessarily involve treating harms beyond those that exist today."

Two further reductions apply. The court declined to fund construction of new hospitals or clinics, and it reduced allocations to reflect Meta's market share, acknowledging that Snapchat, TikTok, YouTube, X, Reddit and Discord also contribute to the harms at issue. The State's own designee identified those platforms in deposition testimony, and the court recorded that Snapchat is the primary platform used by criminals to carry out sextortion.

Compliance obligations run five years from entry of the order, tolled during any appeal provided Meta posts a supersedeas bond. Meta must file written status reports with the court and the State by June 30 and December 31 of each year during that period.

The factory analogy places advertising inside the nuisance

The most consequential passage for the advertising industry is not in the remedies section. It is in the conclusions of law, where Biedscheid adopted an analogy Meta's own counsel had introduced.

Meta had argued that a court abating pollution from a factory may order the emissions stopped but may not close the plant, obstruct its production function, or make the owner treat unrelated community illness. The court agreed with the framework, then applied it: "The Court considers Meta's platforms to be analogous to the factory, the advertising and other content displayed on those platforms to be what is produced by the factory, and the psychological harm to and sexual exploitation of children to be the pollution that must be abated."

Advertising is, in that construction, the output. The pollution is a byproduct of the machinery that delivers it. That framing narrowed what the court would do next, because anything touching content presentation ran into the First Amendment and Section 230 of the Communications Decency Act.

What the court refused

Biedscheid granted no relief at all on algorithm design. The State's proposals were "vague and aspirational, rather than objective and measurable," the order states, and would interfere with Meta's business to a degree the court compared to placing the factory in receivership. Because algorithmic ranking directly alters how content is displayed, the court concluded such orders would likely violate the First Amendment and Section 230.

Infinite scroll and autoplay survived on similar grounds. The order records that the evidence at trial established these features, along with like counts and push notifications, create a product that is "highly rewarding psychologically and neurochemically" for adolescents. It also records that autoplay, infinite scroll and algorithmic recommendation are used widely across the industry, and that restricting only Meta "could harm the viability of Meta and its platforms." Regulation of industry-wide features, the court held, belongs to legislatures and executive agencies.

The State's request for a Child Safety Monitor was denied. Applying the factory metaphor again, Biedscheid wrote that the State asked a monitor not merely to measure pollutants but to decide which pollutants to test, set acceptable levels, and compel the owner to manufacture a different product.

Age verification hit a statutory wall. The court concluded that the Children's Online Privacy Protection Act and the Federal Trade Commission's COPPA Rule prevent it from ordering Meta to request personal data from children or track them passively, even for age checks. A 2026 FTC policy stating it will not enforce COPPA in certain age-assurance contexts does not alter the rule, the order notes, because that policy "does not create any substantive rights or entitlements." Biedscheid also observed that the United Kingdom, European Union and Australian restrictions frequently cited in argument were enacted by parliaments, not courts.

WhatsApp was excluded entirely. The court found it is not a contributing cause of the nuisance because it does not recommend content or people, and left its end-to-end encryption untouched. Facebook encryption also survived. Instagram's cessation of end-to-end encryption, already implemented by Meta, must remain in place for the abatement period.

What Meta must build

The affirmative measures cluster around three areas where the court found it could act without touching content ranking.

Notification blackouts and a monthly ceiling

Meta must eliminate push notifications to known or estimated under-18 accounts in New Mexico between 10:00 PM and 7:00 AM every day, and between 8:00 AM and 3:00 PM on academic-year weekdays. Messages from connected users and urgent security or hazard alerts are carved out.

Like counts must be hidden by default on every under-18 account, overridable only with a parent or guardian's permission.

The order also imposes a hard usage ceiling: no more than 90 hours of use per month, cumulatively across Facebook and Instagram, for accounts belonging to users under 18. That works out to roughly three hours a day. Internal Meta data cited at trial showed the top 1% of New Mexico teens by time spent on Facebook logging at least 290 minutes daily on that platform alone.

Age assurance within COPPA limits

Meta must proactively request proof of age from New Mexico accounts its systems predict belong to under-13 users, and mark for deletion accounts that do not verify within 30 days. It must attempt to develop a dedicated under-13 prediction model within two years, using reasonable best efforts given COPPA constraints. Where Meta cannot estimate a precise age, it must treat the user as under 13 or under 18 by default until verification.

Additional obligations include reviewing the friend networks of deleted underage accounts, partnering with schools or a child safety organisation to build a reporting portal covering any social media platform, reducing the steps required to report an underage account, and maintaining an appeals process.

Teen Accounts must persist through the abatement period, with private-by-default settings, a prohibition on recommending any under-18 account to an unconnected adult, and a bar on unconnected adults messaging minors.

Chatbots and law enforcement reporting

The order singles out artificial intelligence chatbots for a pointed observation. Meta spent considerable time arguing that adolescents circumvent its safety tools, the court noted, then "expressed exasperated surprise" that users were not interacting with chatbots as intended. Meta must now prohibit under-18 New Mexico users from romantic or sexualised interactions with its chatbots, and prohibit adult users from prompting chatbots to simulate such interactions involving a person under 18.

On reporting, testimony from Jonathan Talley, commander of the New Mexico Internet Crimes Against Children division, established that roughly 75% of Meta's CyberTip reports for the state require law enforcement to obtain a warrant before review, a higher rate than any other platform, because Meta forwards reports without human review. A majority, he testified, are "junk tips."

Meta must now ensure a human reviews every CyberTip report of new child sexual abuse material posted by or shared with New Mexico users before referral to the National Center for Missing and Exploited Children. Within 30 days of the order, Meta must contact NCMEC to discuss improvements to the report format and report a baseline CSAM detection rate. By the end of the five-year period, that detection rate must have improved by at least 5%.

Disclosure requirements complete the package. Meta must display a banner screen linking to its Transparency Center, Parents Guide, Safety Center, Help Center and Teen Accounts materials to New Mexico adults at least monthly and to under-18 users at least weekly. New under-18 users get daily information screens for their first 30 days. Both must be submitted to the State for review and approval. Nathaniel Lubin, an expert on platform abatement, had testified that Meta's Community Standards Enforcement Reports measure the company's own detection performance rather than actual prevalence of harm.

The evidence the court accepted

Dr. Zachary Ward's analysis of New Mexico Youth Risk and Resiliency Survey data produced the attributable-fraction figures the court relied on: 1 in 10 children aged 11 to 17 in the state has at least one mental health concern that would not exist absent social media exposure, and among those with a concern, 21% would not have had it. Component estimates ran to 26% for depression symptoms, 20% for anxiety, 15% for sleep deprivation, 22% for at least one suicide risk factor, and 40% for disordered eating among middle schoolers.

Meta's expert Dr. Justin McCrary criticised that analysis. The court found his own presentation of the data incomplete and insufficient to displace Ward's.

Internal Meta documents supplied the intent evidence. A June 2019 internal test reported that "26% of the people we recommend to Groomers were Teens." An October 2020 internal chat recorded that People You May Know "was responsible for 80% of violating adult/minor connections." A May 2020 presentation stated that "Approval and acceptance are huge rewards for teens and interactions are the currency on [Instagram]," and mapped fewer notifications to fewer sessions to less time spent. An August 2022 presentation titled "Late Night Use" found that 18.3% of weekly active Instagram teens exhibited behaviours associated with problematic use.

Dr. Damon McCoy, admitted as an expert in social media integrity systems, testified that autoplay, endless scroll and low-value notifications were regarded inside Meta as "dark patterns and addictive-like designs."

Section 230 rejected again

Meta raised Section 230(c)(1) as an affirmative defence. The court rejected it, finding no binding New Mexico authority but adopting reasoning from the Nevada Supreme Court in TikTok, Inc. v. Eighth Judicial District Court, the Ninth Circuit in Lemmon v. Snap, and the Third Circuit in Anderson v. TikTok. Because the State sought liability for platform design rather than third-party content, the statute does not apply.

That conclusion aligns with the Massachusetts Supreme Judicial Court, which ruled in April 2026 that Section 230 does not shield Meta from state design claims, and with a federal judge in California who on August 5, 2026 found that algorithmic feeds served to minors are not protected speech.

What it means for media buyers

New Mexico is a small market. Meta's internal data showed an average of 180,116 monthly active Instagram users aged 13 to 17 in the state between 2018 and 2024, reaching 195,453 in March 2024 against 1,030,640 adult users. The direct inventory impact is marginal.

The template is not. A 90-hour monthly cap, notification curfews and default-hidden like counts are measurable, auditable obligations that any other state attorney general can now cite as precedent. The court's refusal to touch ranking preserves the delivery mechanism advertisers actually buy, but the reasoning behind that refusal was explicitly jurisdictional and constitutional rather than substantive. Biedscheid agreed with the State that these features facilitate problematic use, then said the remedy belongs elsewhere.

Elsewhere is already moving. The European Commission issued preliminary findings on July 10, 2026 that Instagram and Facebook breach the Digital Services Act through addictive design, naming infinite scroll, autoplay, push notifications and recommender systems, with exposure up to 6% of global turnover. Australia's under-16 ban took effect on December 10, 2025, and France approved an under-15 ban in July 2026.

Cost is becoming visible in the accounts. Meta reported second-quarter 2026 advertising revenue of $59.36 billion on July 29, alongside a $2.4 billion legal charge that cut net income 8% to $15.8 billion and pushed operating margin from a long-held 40% to 48% band down to 31%. The New Mexico judgment adds $942 million in combined penalty and abatement obligations to that column, before appeal.

The order anticipates that appeals will follow. It provides for tolling of the five-year abatement period during their pendency, conditional on Meta posting a supersedeas bond, which means the compliance clock may not run for some time. Biedscheid also signalled that further orders will govern administration of the fund and the drawdown process, with additional briefing and hearings to come.

Timeline

Summary

Who: Meta Platforms, Inc., defendant; the State of New Mexico acting through Attorney General Raul Torrez, plaintiff; Chief District Court Judge Bryan Biedscheid, Division VI of the First Judicial District Court. Expert testimony came from Dr. Zachary Ward, Dr. Dana Weiner, Dr. Kelly O'Donnelly, Dr. Jean Twenge, Dr. Kimothi Cain, Dr. Damon McCoy, Dr. Arvind Narayanan, Nathaniel Lubin and others, alongside New Mexico educators and law enforcement officials.

What: A judgment finding that Instagram and Facebook constitute a statutory and common law public nuisance in New Mexico, ordering a $567,000,000 abatement fund on top of a $375,000,000 jury penalty, and imposing five years of specific product and reporting obligations covering age assurance, teen account defaults, push notifications, like counts, a 90-hour monthly usage cap, chatbot interactions, CyberTip report handling and platform disclosures. The court denied all requested relief concerning recommendation algorithms, infinite scroll, autoplay and a child safety monitor.

When: Filed August 6, 2026, following a jury trial held February 2 to March 24, 2026 and a bench trial held May 4 to May 22, 2026. The original complaint was filed December 5, 2023. The abatement period runs five years from entry, tolled during appeal on posting of a supersedeas bond.

Where: First Judicial District Court, Santa Fe County, New Mexico, case number D-101-CV-2023-02838. The abatement measures apply to New Mexico accounts on Facebook and Instagram. WhatsApp is excluded.

Why: The court found that Meta's engagement-optimised design substantially contributed to a youth mental health crisis and to child sexual exploitation in New Mexico, and that those harms migrate beyond the platforms into schools, hospitals and law enforcement systems, interfering with rights common to the public. Section 230 was held inapplicable because the claim targets platform design rather than third-party content.