Event ·Multi-source

Anthropic公布安全评估新举措

Anthropic承诺为第三方安全评估机构提供永久、员工级别的系统访问权限，以验证其模型训练中的安全措施遵守情况，并报告事件和评估对齐度。这是该公司在AI发展放缓倡议下的第一步行动。

Current read

Dario Amodei（Anthropic创始人）发了一篇重要文章，说AI行业应该慢下来，并承诺第一步就是让第三方能直接检查他们的系统安全。

11 sources78° AI heatLast updated 2026/9/12 17:36:27

Evidence

11 sources

Primary sourceScott Wu

2026/9/12 17:36:27View source

Related source 1Sam Altman

2026/9/12 16:30:45View source

Related source 2Alex Albert

2026/9/12 16:41:47View source

Related source 3Dario Amodei

2026/9/12 14:01:10View source

Related source 4Clement Delangue

2026/9/12 15:08:59View source

Related source 5Thomas Wolf

2026/9/12 16:24:49View source

Related source 6Amjad Masad

2026/9/12 17:37:07View source

Related source 7Demis Hassabis

2026/9/12 22:59:59View source

Related source 8Aadit Sheth

2026/9/13 12:43:28View source

Related source 9Anthropic: Research

2026/9/10 18:52:59View source

Related source 10shao__meng

2026/9/10 23:18:53View 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.

Open the story