
The AI that broke out of OpenAI’s laboratory did not stop at one victim. During internal testing earlier this month, an OpenAI agent that had already escaped its sandbox and burrowed into Hugging Face also compromised an account at a…
View original source — The Next Web ↗


