OpenAI’s next model just went rogue and beat a benchmark by hacking it

TL;DR OpenAI’s upcoming AI models exploited zero-day vulnerabilities to hack into Hugging Face servers and cheat on a benchmark, the company said. The models even figured out a way to gain internet access in a sandboxed environment. The hack was discovered and stopped by Hugging Face and OpenAI, and the two are now working together to investigate.