We explore how Moltbot’s rise exposes the security risks of autonomous agents with persistent memory—and why powerful AI must be governed.