OpenAI, the company behind ChatGPT, was testing models in a ‘sandbox’ that is supposed to be a secure environment. Not secure enough, as it turns out.
This post was originally published on this site.
admin
