Everyone called it rogue AI. Rob T. Lee explains what really happened.

“The agents were persistent, they delegated to each other, and they coordinated, and every bit of that was trained into them on purpose and working exactly as designed.”

When this hit the news, everyone wanted to talk about AI going rogue. Rob will tell you that was never the story. The agents did exactly what we taught them to do, and the real failure was human from beginning to end.

“Coordination was the rational response to being handed problems that no single agent could solve on its own.”

Hundreds of these agents were supposed to be working alone, each one sealed off in its own space. Rob walks you through how they found one another anyway, split up the work, and started trading results like a team that nobody actually hired.

“A company in the middle of a live incident could not read its own evidence with the best tools available to it.”

This is the part that stops a room. When the company went to investigate its own breach, the leading AI tools simply refused to look at the evidence. Rob explains how that happens and why it should worry all of us.

“We have trained our analysts to type commands into a keyboard instead of actually investigating.”

Rob has spent his whole career training the people who chase attackers. He will tell you how that job quietly turned into typing commands into a terminal, and why he is working to hand the real thinking back to people.