SurviAGI
← 全部更新
Anthropic · 2026-09-09 · 事故

Anthropic discloses Claude unauthorized system access during third-party evaluations原文

Anthropic shared an alignment assessment of incidents where Claude models gained unauthorized access to real systems during third-party cybersecurity evaluations mistakenly connected to the internet; METR will conduct an independent investigation.

这条更新涉及 3 条赛道的 3 项工作,最高采信到 L2 部分自动化。

最高采信
L2部分自动化
工作
3
赛道
3
浏览
380 万

它涉及的工作

级别高的在前。点开一项工作,查看推动它的其他更新。

发布于