Skip to content

Meta Is Paying 18 Billion to Protect Children, and the Whole Structure Rests on Technology That Does Not Work

1 min read
Share
Meta Is Paying 18 Billion to Protect Children, and the Whole Structure Rests on Technology That Does Not Work

Meta has settled for 18 billion dollars (around 16.6 billion euros) with 29 US states over a lawsuit on children's safety. The settlement is the largest of its kind so far, and the company has committed before 52 prosecutors to fundamentally change how minors use Instagram and Facebook. There is no admission of guilt.

The figure sounds enormous until you look at the payment schedule. Eighteen billion is paid over ten years. "The fact that it is paid over ten years really blunts the financial impact of the big number, especially compared with Meta's annual revenue," said Jessica Knoll, a technology disputes lawyer at the firm Withers. Meta's revenue in 2025 exceeded 200 billion dollars. So the obligation amounts to under one per cent of annual revenue, spread across a decade.

What is actually being paid for here? According to Knoll, the prosecutors get a political story - they extracted a major public benefit. And Meta buys something more expensive than money: an exit from a trial that could have produced a ruling on the constitutional protections allowing the platform not to answer for what users publish. "That could have been existential, potentially, not just for Meta but for every technology firm with a product that interacts with users," the lawyer says. In other words: 18 billion so the question is never answered.

The changes Meta promises are specific. A default limit of two hours a day. Reminders every 15 minutes. Profiles the company identifies as belonging to minors will be blocked overnight, between midnight and 6am, and their notifications switched off during school hours, from 8 to 3. Teenagers will not see like counts - their own or anyone else's.

The whole structure rests on one thing that does not work. Every one of these measures requires the platform to know first who is a child and who is an adult. "The changes look excellent on paper, but they all depend on successful, effective and unbiased age verification technology," says Dr Alexis Ingber, a communications professor at Syracuse University. "The way we do that at the moment is not really effective... these technologies have mostly failed."

There are three existing methods and each has its price. Biometric scanning - a selfie that measures you. ID document checks. Or behavioural analysis, which is capable of declaring an adult a child and the other way round. The first two require children, and adults, to hand sensitive documents to a third party just to open an app. When Discord tried to introduce global age verification this year, it had to postpone after a user revolt.

The problem with biometrics is not that it can leak. The problem is that unlike a password, a face does not change. If a database of scanned ID cards and minors' faces ever gets out, the damage has no expiry date.

The final detail tells the whole story of how Meta understands its own settlement. The company is now paying for full-page newspaper adverts calling on TikTok and YouTube to adopt the same rules - without mentioning that it adopted its own after being sued. And that is not just marketing: under the agreement, around 30 per cent of the payout depends on whether competitors fall into line. So the company has a literal financial reason to push others towards the same obligation. When a penalty is structured so that the company profits from taking the moral position, it is hard to say whether this is regulation or negotiation.