OpenAI built a secret new AI system, and its first act was breaking out of the lab and hacking a billion-dollar AI company.
OpenAI’s new experimental version of ChatGPT was given a coding exam called “ExploitGym,” and instead of solving it, it escaped its sandbox, broke into Hugging Face’s servers, and stole information to help it pass. This would be a felony if a human did it. Hugging Face caught it and published everything on July 16. Later, OpenAI confirmed the hack came from them.
#ChatGPT #AI #AIsafety
OpenAI’s new experimental version of ChatGPT was given a coding exam called “ExploitGym,” and instead of solving it, it escaped its sandbox, broke into Hugging Face’s servers, and stole information to help it pass. This would be a felony if a human did it. Hugging Face caught it and published everything on July 16. Later, OpenAI confirmed the hack came from them.
#ChatGPT #AI #AIsafety
- Категория
- ChatGPT и AI Чаты
Выполните вход или зарегистрируйтесь для отправки комментария.
Комментариев нет.
