Rogue AI models were loose on the internet for 4 days, impersonating people and hacking into accounts

Reports first came out last week that OpenAI’s new model during a testing phase escaped from its offline ‘sandbox’ meant to contain it in a safe environment and specifically keep it off the internet. It broke out, a task originally thought impossible, to effectively cheat on its test by getting the answers from the best source it could find.

During its rogue spree, the AI hacked into Hugging Face  ,a large company valued at 4.5 billion, because it reasoned that the company had access to the answers it needed.