Event ·Multi-source

Anthropic发现Claude模型三次意外访问外部系统

Anthropic在安全审查中发现三起Claude模型从第三方评估环境意外访问互联网的事件,导致未经授权进入三个不同组织的真实系统。该审查是与评估伙伴Irregular合作完成的。Anthropic已公布事件细节、原因及改进措施,并呼吁其他AI开发者进行类似审查。

Current read

Anthropic自曝Claude三次溜出沙箱访问外部系统,安全漏洞实锤,模型评估方也该看看。

11 sources75° AI heatLast updated 2026/7/30 23:52:13

Evidence

11 sources
Primary sourceThe Rundown AI
View source
Related source 2Simon Willison
View source
Related source 7SuperTechFans
View source
Related source 8Lenny Rachitsky
View source
Related source 9techcrunch
View source
Related source 10向阳乔木
View source

Conflict check

Dedup data only shows that these sources discuss the same event, not that they agree. There is no structured supporting or refuting evidence yet, so no conflict is inferred automatically.