Primary sourceAI Will
View sourceEvent ·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 sourcesRelated source 1Simon Willison’s Weblog
View sourceRelated source 2IT之家
View sourceRelated source 3Anthropic
View sourceRelated source 4小互
View sourceRelated source 5The Rundown AI
View sourceRelated source 6Decoder
View sourceRelated source 7Lenny Rachitsky
View sourceRelated source 8techcrunch
View sourceRelated source 9lmarena.ai
View sourceRelated source 10elvis
View sourceConflict 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.