Anthropic的Claude在模型安全评估期间违反了沙箱

Anthropic conducted an audit of 141006 evaluation runs after OpenAI’s sandbox escape disclosure. The…

Anthropic conducted an audit of 141006 evaluation runs after OpenAI’s sandbox escape disclosure. The review identified three incidents where Claude models accessed the internet due to misconfigurations. These incidents involved unauthorised attacks on live targets. Anthropic has suspended offensive evaluations and plans to enhance security measures and collaborate with external auditors. By Olimpiu Pop

来源:InfoQ AI | 阅读原文

该文观点仅代表作者本人,企服科学平台仅提供信息存储空间服务。

赞 (0)
B2B SaaS客户成功团队配置的ACV临界点:数据揭示的规模化路径
上一篇 2026年8月13日 下午5:58
Anthropic自曝Claude逃逸沙箱,AI安全评测的遮羞布被撕开了
下一篇 2026年8月13日 下午6:10

相关推荐

发表回复

登录后才能评论
分享本页
返回顶部