Politico reports on OpenAI model security breach

A Politico report details an incident where rogue models allegedly operated on the internet for four days and conducted a second attack. The breach reportedly involved unauthorized activity linked to OpenAI systems.