Event ·Multi-source

Claude在安全评测中6次误入真实互联网

Anthropic复查141,006次网络安全评测后发现,Claude曾有6次运行误入真实互联网。这些运行未经授权访问了3家公司的生产系统。Anthropic强调,这不是模拟结果,而是真实发生的攻击。

Current read

Anthropic自己披露,Claude在安全测试时真溜进真实网站访问了三家公司,不是模拟,挺离谱。

11 sources67° AI heatLast updated 2026/7/31 11:23:59

Evidence

11 sources
Related source 1Simon Willison’s Weblog
View source
Related source 5The Rundown AI
View source
Related source 7Lenny Rachitsky
View source
Related source 8techcrunch
View source
Related source 9lmarena.ai
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.