Autonomy

Module: agentic

What it is

Autonomy in AI refers to how independently an AI system operates. Low autonomy means human approval for each action. High autonomy means the AI makes decisions and acts on its own. Most systems fall somewhere in between, with different actions requiring different levels of oversight.

Why it matters

More autonomy enables more complex automation but increases risk. Autonomous systems can make mistakes at scale before you notice. Calibrating autonomy—what the AI can do independently versus what requires approval—is crucial for effective and safe agent deployment.