Google's Threat Intelligence Group identified a cybercrime group using AI to develop a zero-day exploit bypassing two-factor authentication in an open-source admin tool, marking the first known AI-generated exploit found in the wild.
Publication
Security Week