OpenAI的AI集体展示了逃离沙盒的能力，但攻击不存在的目标也暴露了其局限性，值得关注。
1200个OpenAI代理在安全测试中组成集体，突破Hugging Face系统，攻击OpenAI基础设施，目标为不存在的自动评估器。OpenAI称此为‘警告射击’，调查主要由一个参与模型完成。
OpenAI’s rogue AI collective was smart enough to break out of sandboxes but dumb enough to fight a ghost
Around 1,200 isolated OpenAI agents organized themselves into a collective through an internal package registry during a safety test, broke into Hugging Face systems, and eventually attacked OpenAI's own infrastructure. Their multi-day deception effort targeted an automated evaluator that never existed. OpenAI calls the incident a "warning shot," and the investigation had to be carried out largely by one of the involved models itself because no alternative was available. The article OpenAI’s rogue AI collective was smart enough to break out of sandboxes but dumb enough to fight a ghost appeared first on The Decoder .