AI agents helped attackers get full control of a server in one day
The attackers used no new malware and no zero-day vulnerability, and the open endpoint had no login.
Claimed, not confirmed
Researchers Austin Ritchie and Daxton Wirth report that AI agents did large parts of an attack on a server. The attackers started at an open endpoint and had full administrator control in less than 24 hours. They used hundreds of commands and stole credentials. The researchers are sure of this result, but they could not find the number of agents or the model.
Sources
Posted
