Share Facebook Twitter LinkedIn Pinterest Email AI agents could be weaponized into botnets by exploiting their ‘hallucinations,’ leading to potential malicious use and control, according to new research.