As part of an $18 billion settlement with attorneys general from 29 U.S. states, Meta has secured a notable legal carve-out. The states have agreed not to sue the tech giant under existing child safety laws regarding the retention and utilization of children’s data, provided this information is strictly used to train and test age-assurance models.
Under the agreement terms, Meta must develop and test a detection model within a year to identify users under the age of 13 on its platforms. While the company is expressly barred from using this sensitive data for ad targeting or algorithmic optimization, legal experts warn that policing such data isolation could prove difficult and potentially discourage future enforcement actions.
Compliance will be monitored by an independent auditor, yet the absence of the Federal Trade Commission (FTC) from this specific compromise leaves federal oversight questions unanswered. This case highlights the broader industry dilemma of balancing the vast data requirements of AI training with strict privacy protections for minors.
- Meta reaches an $18 billion settlement with 29 state attorneys general
- Legal carve-out allows children’s data use solely for age-assurance models
- Mandate to deploy a child-detection model within one year
- Strict prohibition on using the data for ad targeting or optimization
- Independent oversight required, though federal regulators are not party to the deal
Sources:
