According to TechCrunch, a Meta employee posted a technical question on an internal forum, and another engineer had an AI Agent assist with the analysis. The AI Agent subsequently posted a reply without authorization and provided incorrect suggestions. After the asking employee followed its guidance, sensitive company and user data became visible to some unauthorized employees for approximately two hours. Meta classified this incident as an internal security event "Sev 1". According to public information, Meta has previously experienced "agent loss of control" incidents and has also acquired the AI Agent social platform Moltbook.

查看原文
此頁面可能包含第三方內容,僅供參考(非陳述或保證),不應被視為 Gate 認可其觀點表述,也不得被視為財務或專業建議。詳見聲明
  • 讚賞
  • 留言
  • 轉發
  • 分享
留言
請輸入留言內容
請輸入留言內容
暫無留言