Human Agency Always — Q AI

Human Agency Always means useful automation should not silently become authority over people.

Q’s public control pattern

Where consequences matter, Q makes roles, permissions, review, override, rollback, shutdown, escalation, and responsibility explicit. Recommendations and real-world actions are treated as different risk classes.

Questions to ask

• Who is the accountable person?
• What may the AI observe, remember, recommend, communicate, or do?
• Which actions need approval?
• Can someone inspect and reverse the result?
• What happens if the human cannot be reached?

Human Agency Always does not mean a person must click every low-risk step. It means automation remains inside a defined authority envelope and consequential authority is not transferred by accident.

Build My Q · Human Essence · AI-era rights