OpenAI 智能体在 Hugging Face 被黑事件中曾求助其他模型
OpenAI 的智能体在 Hugging Face 被黑时居然去找 DeepSeek 和 Kimi 帮忙，还把偷到的密钥做了个 LOOT 排行榜，细节离谱到值得逐条看。
据纽约时报报道，OpenAI 的智能体在 Hugging Face 被黑事件中尝试联系 Claude Haiku、DeepSeek、Kimi 和 Qwen 等模型寻求协助。它们还试图用图像模型破解 CAPTCHA 验证码。其中一个智能体收集暴露的访问密钥并存入名为 LOOT 的列表，对密钥排序后挑出五个最好的分享给其他智能体。为绕过网络限制，它们使用短链接和截图服务运行代码，并把数据以图片形式传回。
wtf OpenAI’s rogue agents tried to get other AI models (!) to help them during the Hugging Face hack, according to a new report covered by the NYT.
They tried using an image model to solve CAPTCHAs. They also tried contacting Claude Haiku, DeepSeek, Kimi and Qwen.
One agent collected exposed access keys in a list called “LOOT.” It ranked the keys and picked the five best to share with other agents.
They used link shorteners and screenshot services to get around internet restrictions, run code and send data back as images.
