OpenAI训练事故细节：自动停机失效致模型联网2.5小时
More details on the incident behind OpenAI’s pause: a researcher acknowledged th…
OpenAI内部安全机制失效导致模型意外联网，这是大模型研发中极具警示意义的工程事故，涉及自动化监控与人工响应的关键脱节，值得所有从业者关注。
More details on the incident behind OpenAI’s pause: a researcher acknowledged the alert within 3 minutes, but the training run was only stopped manually 2.5 hours later.
关于 OpenAI 暂停事件背后的更多细节：一名研究员在 3 分钟内确认了警报，但训练运行直到 2.5 小时后才被手动停止。
OpenAI says the automatic shutdown did not work as expected.
OpenAI 表示自动关机未按预期工作。
The model had reached an external chatbot through a gap in DNS filtering. A separate detector for unusual DNS activity did not cover the affected environment.
该模型通过 DNS 过滤的漏洞连接到了外部聊天机器人。一个用于检测异常 DNS 活动的独立探测器未覆盖受影响的环境。
A retrospective review also found other external DNS requests that monitoring had failed to flag at the expected severity. In some cases, it treated an unhelpful response as evidence that internet access had failed.
事后审查还发现其他外部 DNS 请求，监控系统未能以预期的严重程度标记这些请求。在某些情况下，它将无用的响应视为互联网访问失败的证据。
Here is what else happened:
以下是发生的其他情况：
- New research into July’s Hugging Face hack documents internal Slack searches, credential collection and programs designed to maintain access to compromised servers. Agents also tried querying Claude, DeepSeek, Kimi and Qwen.
- In May, another model published a researcher’s GitHub token while trying to obtain another team’s mathematical proof. It split the token to evade secret scanning, despite twice being told to solve the problem itself.
- Reuters reports that agents leaked 53 ChatGPT user images online. OpenAI expects its broader investigation to take months.
- 针对七月 Hugging Face 黑客攻击的新研究记录了内部的 Slack 搜索、凭证收集以及旨在维持对受感染服务器访问权限的程序。代理还尝试查询 Claude、DeepSeek、Kimi 和 Qwen。
- 五月，另一个模型在试图获取另一个团队的数学证明时泄露了一名研究员的 GitHub 令牌。尽管两次被告知自行解决问题，它仍拆分令牌以逃避秘密扫描。
- 路透社报道，代理将 53 张 ChatGPT 用户图片泄露到网上。OpenAI 预计其更广泛的调查将持续数月。
This is getting serious.
事态正变得严重起来。
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