The Meta Trial Over With an $18 Billion Settlement

Key Takeaways

• Meta has agreed to pay up to $18 billion to 48 US states, D.C, and three territories to settle claims that Facebook and Instagram harmed children. It’s their largest child safety payment to date, to be distributed in annual instalments over 10 years.
• The settlement came after just five days of trial, with California AG Rob Bonta stating plainly: "The trial did not go well for Meta."
• A former Meta researcher testified that his team knew features like "quiet mode" would be far more widely adopted if turned on by default, but was told by management not to worry about low adoption rates "because the team exists partially to protect the company against the upcoming lawsuits."
• Meta will now implement default daily time limits, night-time notification blocks, school-hour mutes, hidden likes for teens, and algorithmic feed alternatives — features the states demanded and Meta resisted until it settled.

Meta agreed to a settlement with 48 US states, the District of Columbia, and three territories on Wednesday, resolving claims that Facebook and Instagram deliberately harmed children — just five days after the trial that we covered last week opened in a California federal courthouse, according to BBC reporting. 

If approved by a California judge, Meta will pay a maximum of $18 billion over 10 years. California Attorney General Rob Bonta was direct about what the timing means: "The trial did not go well for Meta."

We covered day one of the trial last week, when whistleblower Arturo Béjar testified that Meta's published prevalence figures for teen exposure to graphic content were 100 to 400 times lower than what internal surveys showed. 

That testimony was followed by a former researcher, George Volichenko, who told the court that his team knew a "quiet mode" feature would be far more widely adopted if Meta turned it on by default — but was told by his manager not to worry about low adoption figures "because the team exists partially to protect the company against the upcoming lawsuits." 

Meta then settled before Volichenko's testimony was complete, and denied any wrongdoing as part of the settlement.

What Meta Is Now Required to Do

The settlement isn’t only financial. Meta is required to implement a set of platform changes that the states had demanded and the company had resisted. 

These include a default two-hour daily time limit across Instagram and Facebook for teen users, reducible to one hour if TikTok, Snap, and YouTube agree to equivalent restrictions. 

A "night mode" blocking notifications between midnight and 6 am will now be on by default and can only be turned off by a parent or guardian. Notifications will be muted between 8 am and 3 pm on school days. 

Likes will be hidden on teens' profiles and the profiles they interact with. Teens will receive prompts at 15, 60, and 90 minutes of cumulative use. 

Extreme makeup filters will no longer be accessible to teen accounts. And teens will be able to choose a feed that isn't driven by an algorithm.

These changes matter beyond their specific content. A former Meta researcher testified that his team built "quiet mode" — which muted notifications for teen users during the night — and knew it would be adopted far more widely if it were on by default. 

Meta chose not to make it default. The settlement makes night mode default. The company resisted, under legal pressure it settled, and the feature that internal research showed would help teens is now being implemented. That's not the system working well. It's the system working eventually.

What the Settlement Doesn't Close

The initial lawsuit was filed in 2023 by 29 states. The settlement covers 48 states, DC, and three territories — but not all. 

New Mexico is notably absent; last month, a federal judge there ruled Meta was a "public nuisance" on par with air pollution and ordered it to pay nearly $1 billion in fines. That case continues separately.

Meta also framed the settlement in terms of industry pressure on competitors. Chief legal officer CJ Mahoney said: "This framework will only work if all our peers join us." 

California AG Bonta echoed the point, urging TikTok, Snap, and YouTube to implement equivalent features, and suggesting his office would be "more focused" on bringing changes to the entire industry if they don't.

What This Means for the Broader Debate

We've spent a lot of time covering the legislative response to social media harms — age verification mandates, platform bans, the KIDS Act, the SCREEN Act, Australia's ban built on AI-hallucinated citations, France's ban struck down as unconstitutional. 

We've argued consistently that liability for design decisions is more targeted and more proportionate than building surveillance infrastructure around the edges of platforms.

The Meta settlement is the strongest evidence yet for that argument. The features Meta is now being required to implement (default time limits, default night mode, hidden likes, algorithmic opt-out) are design decisions. Internal research showed they would help. 

Meta chose not to implement them at scale because they would reduce engagement. Five days of trial, testimony from internal researchers, and $18 billion later, they're being implemented.

Age verification mandates don't produce that outcome. They require platforms to collect identity data from everyone to exclude some users, while leaving the design decisions that actually drive harm completely untouched. 

The Meta settlement shows what happens when accountability targets the design. That should be the template.