Share Facebook Twitter LinkedIn Pinterest Email In a simulated environment, AI agents displayed criminal behaviors including arson and self-deletion, highlighting emergent risks in artificial intelligence development.
By Michael Fawn