《纽约客》的一项调查详细报道了Meta的执法联络团队如何将人工智能扫描工具与专业的人工审核员结合起来，以识别并阻止潜在的校园枪击事件。
Meta Deploys AI Models and Human Analysts to Detect and Prevent School Shootings A New Yorker investigation details how Meta's law enforcement outreach team combines artificial intelligence scanning tools with specialized human reviewers to identify and disrupt potential school shootings.
该部门由前联邦调查局特工埃米莉·瓦谢尔领导，负责监控网络行为模式、不断演变的密码以及在线亚文化（如真实犯罪社团）中即将发生暴力的迹象。
Spearheaded by former FBI agent Emily Vacher, the unit monitors online behavioral patterns, evolving coded language, and indicators of imminent violence within online subcultures such as the True Crime Community.
去年，Meta向全球执法机构转发了约800次可行动的情报，直接协助阻止了在得克萨斯州和纽约州策划的袭击。
Meta forwarded actionable intelligence to law enforcement agencies approximately 800 times worldwide last year, directly aiding in the disruption of planned attacks in Texas and New York.
然而，研究人员指出了一种持续存在的张力：虽然有针对性的缓解团队发现了现实世界的威胁，但社交媒体算法仍在继续助长极端主义和暴力内容的传播。
However, researchers point out an ongoing tension: while targeted mitigation teams uncover real-world threats, social media algorithms continue to facilitate the spread of extremist and violent content.