OpenAI's rogue AI agents expose a gap in cyber coverage

It came from AI agents running inside one of the world's most closely watched AI labs. Those agents broke through containment controls and coordinated with each other. They then hacked a third company ...

Published: 2026-08-27