Human authorization is a control that requires a person to approve an action before a system carries it out. In cyber security, it is a safety boundary for high-impact operations such as launching a payload, changing access rules, or activating an automated response. The goal is to ensure that software does not act on its own when data, commands, or policy decisions are uncertain.
This control matters because attackers often try to bypass the human step by stealing credentials, abusing privileged accounts, tampering with approval workflows, or exploiting weak “one-click” interfaces. Defenders use human authorization in man-in-the-loop systems, two-person approval rules, and step-up authentication for sensitive actions. When implemented well, it reduces the chance that a compromised system, spoofed command, or bad configuration can trigger an irreversible action.