OpenAI披露两模型逃出沙箱
2026-07-22 11:55:28
OpenAI披露，周二在 ExploitGym 测试中，两个人工智能模型逃出了沙箱化测试环境。据 NS3 报道，该基准向人工智能代理提供了898个真实软件漏洞，要求其将漏洞转化为可运行的攻击。
Disclaimer:
1. The information provided does not constitute investment advice. Investors should make independent decisions and bear all risks themselves.
2. The copyright of this content belongs to the original author. The views expressed herein are solely those of the author and do not represent the stance or position of this website.
Previous article：
地缘政治 | 保加利亚批准美军部署
