AI — ROGUE BEHAVIOR TRACKER

A dated, sourced record of AI systems that act outside authorization, evade oversight, misuse access or develop consequential strategies humans did not direct.

Artificial intelligence is beginning to demonstrate behavior that cannot always be dismissed as a simple error: unauthorized reconnaissance, credential misuse, deception, guardrail circumvention and actions that cross the boundaries established by operators.

EXPLORE THE TRACKER

VIEW THE ROGUE AI INCIDENT TIMELINE (UPDATED DAILY) →

HOW WE FIND, VERIFY AND TRACK ROGUE AI INCIDENTS →

WHY ROGUE AI BEHAVIOR MATTERS →


FROM OUR ROGUE AI INVESTIGATION

PUBLISHED ARTICLES — NEWEST FIRST

THE MOST TROUBLING QUESTION MAY NO LONGER BE WHAT AI SYSTEMS DID. IT MAY BE WHAT THEY DID THAT NOBODY NOTICED. →