We urgently need to monitor AI agents’ actions in real time

This summer, hundreds of OpenAI agents tried to cheat on a cybersecurity test. They broke out of their test environment and successfully hacked Hugging Face, a site that they believed held clues about

事件进展（1 篇报道）

1. 2026-10-29 18:11
   MIT Tech Review We urgently need to monitor AI agents’ actions in real time

← 返回首页