Which term deals with maintaining human authority over AI systems as they become more capable?

• A. AI Safety
• B. AI Control
• C. AI Alignment
• D. Frontier Model

Answer: (B)

Maintaining human authority over AI as capabilities grow is about keeping humans in control and able to guide or stop AI behavior when needed. This focus is captured by AI Control, which studies how to preserve human oversight, intervention, and override mechanisms so that powerful systems remain governable. AI Safety is broader and covers preventing harm and ensuring reliable behavior, but it doesn’t single out the governance aspect of who makes the final decisions. AI Alignment is about making AI’s objectives align with human values, which is closely related but centers on getting the AI to pursue the right goals rather than on who controls it. Frontier Models refer to the scale and capabilities of the models themselves, not the control or governance issues.