Meta 因安全配置错误导致超过 45,000 张数据库表泄露,于 2026 年 6 月 22 23 日暂停了其“模型能力倡议”(MCI)。该事件包含员工私人聊天记录、键盘和鼠标操作数据等,被内部定为 SEV 2 级严重事故。
MCI 计划于 2026 年 4 月启动,强制收集美国员工的鼠标移动、点击、键盘输入和屏幕截图,用于训练 AI 代理模仿白领工作流程。
尽管内部超过 1,500 名员工曾签署请愿书反对,但最终是这次内部数据泄露——而非长期的反抗——迫使 Meta 暂停了该计划。
Search & fact-check with cited sources for What led to Meta's suspension of its Model Capability Initiative (MCI) after an internal data leaMeta suspended its employee AI training program after a security misclassification exposed sensitive data across the company.
AI 提示
Create a landscape editorial hero image for this Studio Global article: Search & fact-check with cited sources for What led to Meta's suspension of its Model Capability Initiative (MCI) after an internal data lea. Article summary: MCI was suspended not because of the employee backlash alone, but because a **security configuration failure** exposed the program's deeply sensitive data trove to the entire company. The incident turned months of intern. Topic tags: general, general web, user generated, news. Style: premium digital editorial illustration, source-backed research mood, clean composition, high detail, modern web publication hero. Use reference image context only for broad subject, composition, and topical grounding; do not copy the exact image. Avoid: logos, brand marks, copyrighted characters, real person likenesses, fake screenshots, UI text, readable text, watermarks, charts w
openai.com
Meta 在 2026 年 6 月 22 日至 23 日暂停了其“模型能力倡议”(Model Capability Initiative,简称 MCI)。这一决定的直接原因是:一次安全配置错误,导致海量敏感员工数据被整个公司员工访问。该事件在 Meta 内部被定性为 SEV 2(严重性仅次于 SEV 0 和 SEV 1 的第二高等级)。持续数月的内部抗议未能改变局面,但这次数据泄露最终迫使 Meta 采取行动。