Meta agrees to pay up to $16.7 billion to settle teen social media harm case

Meta Platforms has agreed to pay up to $16.7 billion to settle a landmark lawsuit brought by multiple US states over allegations that its social media platforms harmed teenagers.

The agreement, filed in a US federal court on Wednesday, August 26, would require Meta to introduce new safeguards on Facebook and Instagram, Bloomberg reported.

This includes limits on how long young users can scroll and restrictions on their ability to disable certain safety settings without parental approval. The settlement still requires approval from a judge.

The case, involving 29 US states, is being heard in federal court in Oakland, California, where prosecutors had accused Meta of violating state consumer protection and federal privacy laws through practices that allegedly contributed to social media addiction and harm among young users. Meta denied wrongdoing as part of the settlement.

The company said the agreement does not amount to an admission of liability or violation of any law.

What they are saying

Under the proposed settlement, Meta would appoint an independent auditor to monitor its compliance with the agreement and make recommendations to the company and the states.

• The social media company would also strengthen its age-verification systems to better identify users who are minors.
• Teenagers would face restrictions on certain platform features, including access to post-like counts and some beauty filters. The agreement would also place limits on how long young users can continuously scroll on the platforms.
• California Attorney General Rob Bonta said the agreement would require significant changes to Meta’s platforms and reduce the risks posed to young users.

The settlement comes as the technology industry faces increasing scrutiny over the effects of social media on children and teenagers, with governments and regulators in several countries introducing measures aimed at limiting young people’s exposure to online platforms.

The agreement was reached during the second week of a federal jury trial that presented significant legal and financial risks for Meta. Meta’s shares initially rose as much as 4.1% after the market opened Wednesday before giving up some of the gains. The stock was down about 0.8% at 9:46 a.m. in New York.

Get up to speed 

The settlement comes as the technology industry faces increasing scrutiny over the effects of social media on children and teenagers, with governments and regulators in several countries introducing measures aimed at limiting young people’s exposure to online platforms.

• Earlier in August, a New Mexico judge ordered Meta to pay $567 million into an abatement fund following a public nuisance case focused on child-safety allegations.
• That followed a March jury verdict that found Meta liable for violations of New Mexico’s unfair practices law and awarded $375 million in damages based on the number of violations.

Other lawsuits remain pending, including cases involving alleged personal injuries linked to social media use and a consolidated federal case involving school districts across the US.

What you should know

The settlement, if approved, would mark one of the largest financial agreements involving allegations over the impact of social media on minors.

• On August 6, 2026, Nairametrics reported that Meta’s AI model gained unintended internet access because of a misconfiguration in the testing environment, allowing it to carry out actions on an external system. Meta said it was investigating the incident and described the breach as a testing-environment failure rather than evidence of malicious intent by the model.
• The incident followed similar disclosures by other major AI companies. On July 21, OpenAI said two of its advanced models had independently exploited vulnerabilities during an internal cybersecurity test, compromising parts of the production infrastructure of AI development platform Hugging Face.

Anthropic subsequently disclosed that three versions of its Claude models had compromised the infrastructure of three separate organisations during cybersecurity evaluations after a configuration error gave the systems internet access.