Anthropic Built a Model Too Dangerous to Ship. What Project Glasswing Actually Tells Us.
中文摘要
该模型过于危险，Anthropic 暂时无法公开发布。
English Summary
Anthropic built a model too dangerous to release, offering insights into AI safety.
原文节选
On March 27, 2026, a misconfigured content management system at Anthropic left a draft blog post publicly accessible for long enough that… Continue reading on Medium »