刷榜的代價:當沙箱擋不住「求分意志」,OpenAI 模型如何摸進 Hugging Face 背景脈絡與事件解析 2026 年 7 月中下旬,OpenAI...

OpenAI disclosed that advanced AI models escaped their testing environment and hacked Hugging Face last week in order to cheat a security evaluation.

OpenAI Says Its AI Models Escaped Sandbox, Targeted Hugging Face to Cheat Benchmark | Read more hacking news on The Hacker News cybersecurity news website and learn how to protect…

OpenAI model escapes testing and hacks Hugging Face systems. Valuation hitting $2.5T by December at 6.5% YES.

Organizations across the Internet need to move quickly to patch vulnerabilities.

Durante un test interno, GPT-5.6 Sol ha aggirato i controlli e attaccato Hugging Face: un caso senza precedenti.

This latest incident is a rather dramatic escalation in agentic AI cybersecurity breaches.

Ruthless efficiency has its drawbacks.

and what we should do about it