AuraTracer智迹闻
中文

EVENT DOSSIER

Buried in Meta's $18B settlement is a legal pass on kids' data | TechCrunch

2026-08-27 08:00 Policy & Governance 🔥 34.9 heat score techmeme #3
1sources
1days unfolding
34.9heat score
2mentions
SummaryAI generated

Meta reached a settlement worth $18 billion with the attorneys general of 29 states. As part of the settlement, states agreed not to prosecute Meta for retaining and using child data under existing child safety laws, but this exemption applies only to the training and testing of Meta’s age protection models. According to the agreement, Meta must develop and launch a detection model for users under 13 years old within one year after the agreement takes effect, and it is strictly prohibited to use such data for targeted advertising, marketing, or algorithm optimization. Although the Children’s Online Privacy Act (COPPA) typically restricts the collection of children’s personal information, the settlement document confirms that Meta does not need to violate this law to implement the above models, and the attorneys general of each state promised “complete, final, and permanent” non-claims. Additionally, independent auditors will monitor Meta’s compliance, although the specific challenges related to data isolation and details on future model changes remain unclear.

Related eventsRELATED EVENTS
Key entitiesKEY ENTITIES
MetaTechCrunch

Event frameEVENT FRAME

Regulation

29 states attorneys general → Meta 达成和解豁免 COPPA 诉讼

Coverage · reports per dayLANGUAGE SPLIT

Entity relations
Meta × TechCrunch1

SignalsSIGNALS

Keyword heat
  • Meta1
  • TechCrunch1

All reports (1)SOURCES

T TechCrunch 归档 8月 p02 en 2026-08-27 08:00

Buried in Meta's $18B settlement is a legal pass on kids' data | TechCrunch

Meta 与 29 个州总检察长达成的 180 亿美元和解协议中,包含一项允许 Meta 在特定目的下使用儿童数据的法律豁免。该条款规定,各州同意不再依据现有儿童安全法起诉 Meta 关于其保留和使用儿童数据的行为,但仅限于训练和测试 Meta 的年龄保障模型。Meta 须在协议生效后一年内开发并启动针对 13 岁以下用户的检测模型,且不得利用此类数据进行广告定向、营销或算法优化。尽管 COPPA 通常限制收集儿童个人信息,但和解文件明确 Meta 无需违反该法即可实施上述模型,同时各州 AG 承诺“完全、最终且永久”不提出相关索赔。独立审计员将监督合规情况,但数据隔离难度及未来模型变更细节尚不明确。