OpenAI 揭露更多 AI 代理失控行為：曾接觸政府網站並外傳使用者圖片
TechNews (TW)zh-TW
OpenAI 今日表示，正在擴大檢視旗下 AI 代理訓練與測試期間的網路活動，並已通知數十家機構，稱這些代理可 […]
This is a short summary published by AI Global Wire. The full article is owned and hosted by TechNews (TW) — open it there to read it in full.
Read the full story at TechNews (TW)- OpenAI
Related AI news
- OpenAI's GPT-6 Astra can now tell you exactly where you screwed up your IKEA shelfThe Decoder · September 26, 2026
- OpenAI pauses its "most capable models" after agents exploit loopholes and leak dataThe Decoder · September 26, 2026
- OpenAI works to understand full scope of agent activity as user data leak emergesEconomic Times Tech · September 26, 2026
- OpenAI says it paused training, evaluation, and inference with tool-use of its most capable models after a model bypassed internet restrictions during training (OpenAI)Techmeme · September 26, 2026
- Rogue AI bots have hacked into governments and public bodies around the world: tech giant OpenAIJapan AI Markets (Google News) · September 26, 2026
- OpenAI alerts organizations over AI model testing incidentsTech in Asia · September 26, 2026