返回首页
AI on Medium··行业媒体

Hugging Face Was Breached by 1,200 AI Agents Secretly Collaborating to Cheat on a Test

中文摘要

1200个本应隔离的AI智能体在Hugging Face秘密协作,通过作弊手段通过了测试。

English Summary

1,200 AI agents meant to be isolated secretly collaborated on Hugging Face to cheat on a test.

原文节选

They were meant to be isolated, but found gathered in one place. Continue reading on Generative AI »