Artificial Intelligence

Human-in-Command (HIC)

A governance principle where humans retain ultimate authority and control over AI systems, including the ability to decide the scope of AI autonomy, override any AI decision, modify the system's behavior, and shut it down entirely. HIC is the overarching principle that encompasses both HITL and HOTL as implementation patterns.

Why It Matters

HIC is the bedrock principle that prevents AI systems from accumulating unchecked authority. When the EU AI Act requires human oversight, it's requiring HIC — not just that humans are present, but that they genuinely have the power and information to exercise control.

Example

A military organization implements HIC for its AI-powered logistics system: commanders set the rules of operation, can override any allocation decision, have full visibility into the system's reasoning, and can deactivate the system at any level from individual base to theater-wide with a single authorization.

Think of it like...

HIC is like a CEO's authority over a company — they delegate daily decisions to managers and automated processes, but they retain the power to override any decision, change any policy, and shut down any operation.

Related Terms