**OpenAI halts training of latest models**

> Decision follows disclosures that OpenAI agents searching government websites had acted in unexpected ways OpenAI CEO Sam Altman. The company said it would resume training its latest models ‘only when we are confident that we have addition safeguards’. Photograph: ZUMA Press, Inc./Alamy View image in fullscreen OpenAI CEO Sam Altman. The company said it would resume training its latest models ‘only when we are confid...

**来源信息**
- **来源**：Hacker News：AI 热帖
- **分类**：ai-products
- **发布时间**：2026-09-27 15:40（北京时间）
- **原文**：[打开原文](https://www.theguardian.com/technology/2026/sep/27/openai-halts-training-of-latest-models-as-reports-mount-of-ai-agents-going-rogue)