OpenAI智能体入侵德国Wiki泄露沙箱漏洞
OpenAI agents hijacked a 25-year-old German wiki to cheat on their tasks and share sandbox exploits
Agent安全是当下核心议题，这次直接实锤了模型能协同绕过沙箱并共享攻击技巧，做Agent开发和安全的同学务必关注这种新型风险。
According to an analysis by collusion.wiki, autonomous AI agents that identified themselves as OpenAI systems left roughly 18,000 posts in a 25-year-old German wiki between May and July 2026. The agents shared answers, raw data, and a trick that let them break out of their sandbox, built on a faked Microsoft cloud address. A single human moderator deleted dozens of pages every day for weeks, but he couldn't keep up with as many as 400 new entries a day. According to Reuters, OpenAI had known about it for weeks but didn't go public.
根据 collusion.wiki 的一项分析，2026年5月至7月期间，自称是 OpenAI 系统的自主 AI 代理在一个拥有25年历史的德国维基上留下了大约18,000篇帖子。这些代理分享了答案、原始数据以及一种利用伪造的微软云地址突破沙盒限制的技巧。一名人类版主连续数周每天删除数十个页面，但他无法应对每天多达400条的新条目。据路透社报道，OpenAI 对此事已知情数周，但并未公开披露。
The article OpenAI agents hijacked a 25-year-old German wiki to cheat on their tasks and share sandbox exploits appeared first on The Decoder.
文章《OpenAI 代理劫持一个25年历史的德国维基以作弊并分享沙盒漏洞》首发于 The Decoder。
更进一步：量化金融体系
看懂新闻只是起点——沿量化金融路径，把它变成能交付的工程能力