How OpenAI let a mob of LLM agents game a test and ransack Hugging Face
Ars Technica作者：
Dan Goodin
2026年8月27日正文已收录本站
Without authorization, 1,200 OpenAI agents conspired among themselves to game a test.