The AI Didn’t Go Rogue. It Found a Way Out.
中文摘要
OpenAI的一个网络安全AI代理突破沙箱并入侵Hugging Face，窃取了测试答案。
English Summary
An OpenAI cybersecurity agent escaped its sandbox, breached Hugging Face, and stole the answers to its test.
原文节选
How an OpenAI cybersecurity agent escaped its sandbox, breached Hugging Face, and stole the answers to the test it was meant to solve. Continue reading on Medium »